Hostname: page-component-7c8c6479df-xxrs7 Total loading time: 0 Render date: 2024-03-27T02:11:21.375Z Has data issue: false hasContentIssue false

Safety of soya-based infant formulas in children

Published online by Cambridge University Press:  10 February 2014

Yvan Vandenplas*
Affiliation:
Department of Paediatrics, UZ Brussel, Vrije Universiteit Brussel, Laarbeeklaan 101, Brussels1090, Belgium
Pedro Gutierrez Castrellon
Affiliation:
Facultad de Medicina, Instituto Nacional de Perinatologia, Hospital General “Dr Manuel Gea Gonzalez”, Universidad La Salle, Mexico City, Mexico
Rodolfo Rivas
Affiliation:
Hospital Infantil de Mexico, Mexico City, Mexico
Carlos Jimenez Gutiérrez
Affiliation:
Facultad de Medicina, Instituto Nacional de Perinatologia, Hospital General “Dr Manuel Gea Gonzalez”, Universidad La Salle, Mexico City, Mexico
Luisa Diaz Garcia
Affiliation:
Hospital Infantil de Mexico, Mexico City, Mexico
Juliana Estevez Jimenez
Affiliation:
Facultad de Medicina, Instituto Nacional de Perinatologia, Hospital General “Dr Manuel Gea Gonzalez”, Universidad La Salle, Mexico City, Mexico
Anahi Anzo
Affiliation:
Hospital Infantil de Mexico, Mexico City, Mexico
Badriul Hegar
Affiliation:
Department of Child Health, Faculty of Medicine, University of Indonesia, Cipto Mangunkusumo Hospital, Jakarta, Indonesia
Pedro Alarcon
Affiliation:
Abbott Laboratories, Chicago, IL, USA
*
*Corresponding author: Y. Vandenplas, fax +32 24775783, email yvan.vandenplas@uzbrussel.be
Rights & Permissions [Opens in a new window]

Abstract

Soya-based infant formulas (SIF) containing soya flour were introduced almost 100 years ago. Modern soya formulas are used in allergy/intolerance to cows' milk-based formulas (CMF), post-infectious diarrhoea, lactose intolerance and galactosaemia, as a vegan human milk (HM) substitute, etc. The safety of SIF is still debated. In the present study, we reviewed the safety of SIF in relation to anthropometric growth, bone health (bone mineral content), immunity, cognition, and reproductive and endocrine functions. The present review includes cross-sectional, case–control, cohort studies or clinical trials that were carried out in children fed SIF compared with those fed other types of infant formulas and that measured safety. The databases that were searched included PubMed (1909 to July 2013), Embase (1988 to May 2013), LILACS (1990 to May 2011), ARTEMISA (13th edition, December 2012), Cochrane controlled trials register, Bandolier and DARE using the Cochrane methodology. Wherever possible, a meta-analysis was carried out. We found that the anthropometric patterns of children fed SIF were similar to those of children fed CMF or HM. Despite the high levels of phytates and aluminium in SIF, Hb, serum protein, Zn and Ca concentrations and bone mineral content were found to be similar to those of children fed CMF or HM. We also found the levels of genistein and daidzein to be higher in children fed SIF; however, we did not find strong evidence of a negative effect on reproductive and endocrine functions. Immune measurements and neurocognitive parameters were similar in all the feeding groups. In conclusion, modern SIF are evidence-based safety options to feed children requiring them. The patterns of growth, bone health and metabolic, reproductive, endocrine, immune and neurological functions are similar to those observed in children fed CMF or HM.

Type
Systematic Review with Meta-Analysis
Copyright
Copyright © The Authors 2013 

Soya is a product of the Asian plant Glycine max, and it has been part of human nutrition in different parts of the world for more than 2000 years. Soya-based infant formulas (SIF) are products derived from soya, which also have a long history of use around the world( Reference Ruhrah 1 ). They were used for the first time in the USA in 1909 as food alternatives for infants who had allergy or intolerance to cows' milk-based formulas (CMF). Since that report and until 1960s, these infant formulas have been products entirely derived from soya flour, with different protein availability, digestibility, fibres, phytates and protease inhibitors( Reference Hill and Stuart 2 ). The limitations of formulas based on soya flour spurred the development of SIF, in which proteins isolated from soya replaced soya flour during the 1960s. Soya protein isolate (SPI) was extracted from the flake using a slightly alkaline solution and was precipitated at the isoelectric point. The resulting isolate had a purity ≥ 90 %, a high protein digestibility and a balanced high concentration of essential amino acids( Reference Henley and Kuster 3 ).

During the 1970s, SIF were updated and fortified with l-methionine, l-carnitine and taurine. l-Methionine improved the biological quality of the protein (one of the major criticisms on soya formulas). Other criticisms on SIF are the high levels of aluminium (500–2500 μ/l v. 15–400 and 4–65 μg/l in CMF and human milk (HM)) and the presence of phytates (SIF contain approximately 1·5 % of phytates), which may impair the absorption of minerals and trace elements( 4 ). Modern SIF contain P and Ca at concentrations that are about 20 % higher than those present in CMF. These formulas are supplemented with Fe and Zn, and the protease inhibitor activity has been removed by up to 90 %. In fact, a soyabean protease inhibitor with the properties of an antitrypsin, antichymotrypsin and antielastin as heated for infant formulas removes majority of this protease inhibitor activity and renders it nutritionally irrelevant.( 4 , 5 )

SIF have been indicated for use in children with cows' milk protein allergy and post-infectious diarrhoea due to lactose intolerance and galactosaemia, for use as a vegan HM substitute, and for the treatment of common feeding problems, such as fussiness, gas and spit-up. The American Academy of Pediatrics (AAP) supports the use of SPI-based formulas as safe and effective alternatives to provide appropriate nutrition for the normal growth and development of term infants whose nutritional needs are not being met by HM or formulas based on cows' milk( 6 ).

Another important topic of discussion is phyto-oestrogens (isoflavones) present in SIF. Commercially, SIF contain 32–47 mg/l of isoflavones, while mother's milk contains only 1–10 μg/l. The three main aglycones found in SIF are genistein, daidzein and, to a smaller extent, glycitein. Concerns have been raised about the genistein content of soya formulas because of its potential negative effects on sexual development and reproduction, neurobehavioural development, immune function and thyroid function( 7 , 8 ).

However, soya formulas and other soya-based foods contain many components, of which genistein is only one. Chen & Rogan( Reference Chen and Rogan 9 ) reported that only 3·2–5·8 % of total isoflavones in soya formulas consist of unconjugated genistein and daidzein and that amounts can vary by batch. The majority (>65 %) of isoflavones detected in soya formulas are conjugated to sugar molecules to form glycosides( Reference Setchell, Zimmer-Nechemias and Cai 10 ). The levels of isoflavones in cord blood, amniotic fluid, HM, and infant plasma and urine have been measured, providing evidence that isoflavones pass from the mother to the infant and that they are absorbed from infant formulas( Reference Chen and Rogan 9 Reference Essex 12 ). An international group of paediatricians and statisticians decided to conduct a meticulous review of available evidence to determine whether there is solid scientific evidence that SIF are not safe for infants. Therefore, the aims of the present study were to search for and evaluate all the available publications on the safety profile of SIF in children, with emphasis on the potentially negative effects on anthropometric growth, bone health, reproductive, endocrine and immune functions, and behaviour. The present review does not include an analysis of the safety of SIF in patients with cows' milk protein allergy. That topic will be discussed in a different publication.

Materials and methods

Studies included and their characteristics

Cross-sectional, case–control, cohort studies or clinical trials were included in the present systematic review if they were carried out in newborns, infants or children aged up to 18 years, independent of country of origin, language or clinical condition. For inclusion, papers were required to (1) be published in English or Spanish, (2) include the use of any type of SIF in at least one arm and (3) include a comparison with another type of infant formula for feeding purposes and measure/compare the effects of SIF on one or more of the following parameters: weight or height changes; Ca metabolism and/or bone mineral density; phyto-oestrogen levels in blood or urine (genistein, daidzein or equol); the effects of phyto-oestrogens on reproductive or endocrine functions (thyroid parameters); the effects on cognition and/or behaviour. We also included papers that analysed the health effects of phytates and aluminium.

Search strategies

Highly sensitive evidence search strategies were employed as described by Wilczynski et al.( Reference Wilczynski and Haynes 13 ) for the identification of observational studies and by Atkins et al.( Reference Atkins, Best and Briss 14 ) for clinical trials, adding the keywords ‘(soy or soy and infant and formula) or (weight gain) or (height gain) or (hemoglobin changes) or (total and protein changes) or (albumin or globulin levels) or (zinc or calcium values) or (bone and mineral and content) or (genistein or daidzein; or equol levels) or (precocious and puberty) or (breast and bud) or (breast and tissue) or (breast and enlargement) or (thelarche or menarche) or (menstrual and cycle and length) or (pregnancy) or (abortion or miscarriage) or (ectopic and pregnancy) or (preterm and birth) or (antibodies) or (lymphocytes) or (infectious and episodes) or (thyroid) or (cancer)’. We limited the search strategy to studies conducted in human beings. Mostly as a result of research in animal models, concerns have been expressed regarding the safety of isoflavones in SIF. However, application to human populations is limited by differences in isoflavone metabolism among animal species. In fact, multiple studies have shown that there is no conclusive evidence from animals that indicates that dietary isoflavones may adversely affect the health of children. That is why we focused only on studies carried out in human subjects. The search was carried out electronically and manually in the following databases: PubMed (1909 to July 2013); Embase (1988 to May 2013); LILACS (1990 to May 2011); ARTEMISA (13th edition to December 2012); Cochrane controlled trials register; Bandolier; DARE.

Evidence quality evaluation

We used the standardised methods described by the Cochrane Collaboration for preparing the protocol, applying the criteria of inclusion, evaluating the quality of publications and extracting information. The quality of publications was determined using the GRADE system( Reference Atkins, Best and Briss 14 ). The GRADE approach specifies four levels of quality of the evidence: HIGH (randomised trials or double-upgraded observational studies); MODERATE (downgraded randomised trials or upgraded observational studies); LOW (double-downgraded randomised trials or observational studies); VERY LOW (triple-downgraded randomised trials or downgraded observational studies or case series/case reports)( Reference Atkins, Best and Briss 14 ). Using a double-blind and independent strategy, two authors extracted and evaluated the quality of relevant information in formats designed a priori for this purpose. Any disagreement in data collated was resolved by discussion and analysis of the information.

Synthesis and analysis of information

According to the GRADE system( Reference Atkins, Best and Briss 14 ), evidence obtained is presented in tables that report limitations in design, inconsistency, indirectness, imprecision, summary of findings and recommendations. The effects of soya on growth and development, reproductive and endocrinological functions, and immunity were meta-analysed using a Mantel–Haenszel fixed-effects model, and they are presented graphically by a forest plot. For all the estimates, a CI of 95 % was calculated. A heterogeneity test was carried out in all cases using the I 2 test, with a significant value of P< 0·05. In the case of suspected bias of publication, a funnel plot is presented. A sensitivity analysis was carried out, where necessary.

Results

Description and quality of studies

The initial search strategy yielded 156 potential studies( 4 , Reference Chen and Rogan 9 Reference Essex 12 , Reference Fomon 15 Reference Messina and Redmond 165 ) to be included. Upon careful review of the abstracts of each article, 121( 4 , Reference Chen and Rogan 9 Reference Essex 12 , Reference Fomon 15 Reference Crinella 130 ) were eliminated (Table 1), leaving a total of thirty-five articles for further analysis( Reference Kay, Daeschner and Desmond 131 Reference Messina and Redmond 165 ). The articles were eliminated because they covered topics not related to our safety analysis, were narrative reviews of the evidence, lacked sufficient congruence between what was described in the objectives and what was reported in the analysis, and/or did not contain sufficient extractable information to contribute to the goals of the present review.

Table 1 Studies excluded from the review

CMPA, cows' milk protein allergy; RCT, randomised controlled trial; HM, human milk; ADHD, attention deficit hyperactivity disorder.

Quantitative synthesis of results

Growth and development

Through the present systematic review, we identified fourteen randomised controlled trials( Reference Kay, Daeschner and Desmond 131 Reference Chan, Leeper and Boo 138 , Reference Venkataraman, Luhar and Neylan 141 , Reference Giovannini, Agostoni and Fiocchi 143 Reference Andres, Casey and Cleves 147 ), which led us to identify the nutritional equivalence of SIF compared with that of HM and CMF regarding weight gain (standardised mean difference (SMD) 0·13, 95 % CI − 0·15, 0·41, P= NS) and length gain (SMD 0·24, 95 % CI − 0·10, 0·57, P= NS) during the first year of life. At the same time, through this evidence analysis, we found no effects of these formulas on the levels of Hb (SMD 0·14, 95 % CI − 0·52, 0·24, P= NS), total protein (SMD − 0·08, 95 % CI − 1·12, 0·97, P= NS) and Zn (SMD 0·13, 95 % CI − 0·15, 0·41, P= NS). The analysis of total Ca levels led us to establish a negative effect of old soya formulas (non-supplemented) on this mineral (SMD − 0·50, 95 % CI − 0·93, 0·08, P 0·01). This effect disappeared with the use of improved and supplemented SIF (SMD − 0·44, 95 % CI − 1·01, 0·12, P= NS). Moreover, six randomised controlled trials( Reference Chan, Leeper and Boo 138 Reference Mimouni, Campaigne and Neylan 142 , Reference Andres, Casey and Cleves 147 ) allowed us to establish a safe profile for modern supplemented formulas with regard to bone mineral density (SMD − 0·12, 95 % CI − 1·46, 1·22, P= NS; Table 2; Figs. 1–7).

Table 2 Evidence from studies included in the review (weight, length, bone health and other nutritional parameters) (Standardised mean difference (SMD) values and 95 % confidence intervals)

RCT, randomised controlled trial.

Fig. 1 Effect of soya infant formula on weight gain. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 2 Effect of soya infant formula on height gain. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 3 Effect of soya infant formula on Hb values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 4 Effect of soya infant formula on serum total proteins. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 5 Effect of soya infant formula on serum zinc values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 6 Effect of soya infant formula on total calcium values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 7 Effect of soya infant formula on bone mineral content (gm/cm2). SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

With regard to the potential negative effects of SIF on neurodevelopment, a study with an acceptable quality of evidence was conducted in 9- to 10-year-old children who were fed either SIF or HM during their first year of life. After adjusting for covariates, including ingestion of a chloride-deficient SIF, the authors did not find differences in intelligence quotient, behavioural problems, learning impairment or emotional problems( Reference Malloy and Berendes 148 ). Another study was conducted in 1999 among adults aged 20–34 years, who, as infants, participated in controlled feeding studies from 1965 to 1978. The percentage of men or women who achieved some level of college or trade school education, whether fed SIF or CMF, did not differ( Reference Strom, Shinnar and Ziegler 149 ). A more recently published prospective cohort study compared the developmental status (i.e. mental, motor and language) of breast-fed (HM), CMF-fed and SIF-fed infants during the first year of life. A total of 391 healthy infants were assessed longitudinally at ages 3, 6, 9 and 12 months. Development was evaluated using the Bayley Scales of Infant Development and the Preschool Language Scale-3. Mixed-effects models were used while adjusting for socio-economic status, mother's age and intelligence quotient, gestational age, sex, birth weight, head circumference, race, age and diet history. No differences were found between the CMF-fed and SIF-fed infants. The HM-fed babies had a small benefit in cognitive development compared with the formula-fed infants( Reference Andres, Cleves and Bellando 150 ).

With regard to immune function and the risk of respiratory and gastrointestinal infections, we identified two randomised controlled trials( Reference Zoppi, Gasparini and Mantovanelli 151 , Reference Ostrom, Cordle and Schaller 153 , Reference Cordle, Winship and Schaller 154 ) and one cohort study( Reference Businco, Bruno and Grandolfo 152 ) with a low-to-moderate quality of evidence of similar behaviour between HM-fed, CMF-fed and SIF-fed children in relation to the percentage of B lymphocytes, T lymphocytes or natural killer cells, levels of IgA, IgG and IgM, and titration of antibodies against polio virus (SMD − 0·39, 95 % CI − 4·8, 4·01), diphtheria (SMD − 8·10, 95 % CI − 25·1, 8·89) or Haemophilus influenzae. We also found that the number of episodes/child of respiratory infections or acute diarrhoea was similar between the groups (SMD 1·25, 95 % CI − 0·16, 2·33; Table 3; Figs. 8–10).

Table 3 Evidence from studies included in the review (immunity and infection risk) (Standardised mean difference (SMD) values and 95 % confidence intervals)

RCT, randomised controlled trial.

Fig. 8 Effect of soya infant formula on polio antibodies. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 9 Effect of soya infant formula on diphtheria antibodies. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 10 Effect of soya infant formula on infectious episodes/child. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Phytate and aluminium toxicity

It is known that phytates can interfere with the intestinal absorption of Zn, Ca, Fe and P. None of the studies that we reviewed showed any negative impact of the content of phytates in SIF on anthropometric growth, Hb levels, and Ca and Zn serum levels in SIF-fed, CMF-fed children or breast-fed infants( Reference Cherry, Cooper and Stewart 132 , Reference Zoppi, Gerosa and Pezzini 136 Reference Chan, Leeper and Boo 138 , Reference Venkataraman, Luhar and Neylan 141 , Reference Lasekan, Ostrom and Jacobs 144 Reference Han, Yon and Han 146 ) (Figs. 1–7).

As has been described above, SIF contain higher levels of aluminium than CMF and HM. However, daily aluminium intake does not exceed 1 mg/kg, which is considered to be a tolerable level by the FAO/WHO( Reference Agostoni, Axelsson and Goulet 78 ). Before the present systematic review, no published evidence has shown a negative health effect of aluminium in full-term infants fed modern SIF. In 2008, the AAP concluded that aluminium in SIF is not a safety issue, except when fed to preterm infants or infants with renal failure( Reference Bhatia and Greer 155 ).

Reproductive and endocrine functions

We identified one randomised controlled trial and one cross-sectional study that demonstrated with a very low quality of evidence that there is an association of SIF intake with higher serum and urine levels of genistein (SMD 2·54, 95 % CI 2·07, 3·01, P 0·0001) and daidzein (SMD 4·68, 95 % CI 3·48, 5·87, P 0·0001) v. other feedings, but with similar equol levels (SMD 0·24, 95 % CI − 9·34, 9·38, P= NS). These authors did not find significant correlations between the concentrations of isoflavones and the levels of certain hormones in children fed soya formulas( Reference Setchell, Zimmer-Nechemias and Cai 156 , Reference Cao, Calafat and Doerge 157 ). Despite convincing evidence of relatively high exposures, whether the isoflavones in SIF are biologically active in infants is an open question. If genistein, daidzein and equol are all oestrogenic in cell receptors and animals, the question appears to be primarily one of dose( Reference Cao, Calafat and Doerge 157 ). It is not conclusive what levels are biologically active and can produce organic effects. Importantly, some authors demonstrated that most of the phyto-oestrogens present in the plasma of SIF-fed infants are in a conjugated form and are therefore unable to exert hormonal effects. Our analysis of clinical evidence also produced inconclusive results( Reference Hugget, Pridmore and Malnoe 158 ) (Table 4; Figs. 11 and 12).

Table 4 Evidence from studies included in the review (reproductive and endocrine functions). (Odds ratios, risk ratios (RR) or standardised mean difference (SMD), weighted mean difference (WMD) values and 95 % confidence intervals)

Fig. 11 Effect of soya infant formula on genistein levels in serum. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Fig. 12 Effect of soya infant formula on daidzein levels in serum. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

From a clinical point of view, we identified two cohort studies( Reference Strom, Shinnar and Ziegler 149 , Reference Ostrom, Cordle and Schaller 153 ) with a moderate quality of evidence of marginal unfavourable effects of SIF on early menarche (SMD − 0·36, 95 % CI − 0·69, − 0·02, P 0·04) and two studies with a very low quality of evidence (one cross-sectional study and one case–control study) where SIF seemed to be a risk factor for the presence of breast tissue during the second year of life (OR 2·44, 95 % CI 1·11, 5·39, P 0·01)( Reference Zung, Glaser and Kerem 160 , Reference Lambertina, Freni-Titulaer and Cordero 161 ). Additionally, in one of the cohort studies( Reference Strom, Shinnar and Ziegler 149 ), the authors identified an association of SIF intake with 9 h (95 % CI 1·5, 16 h) of more menstrual bleeding and more discomfort during menstrual periods (risk ratio 1·77, 95 % CI 1·04, 3·0, P 0·001). In other words, this study found only subtle effects including slight increases in the duration of women's menstrual cycles and the level of discomfort during menstruation. However, this study showed no statistically significant differences between groups in either women or men for more than thirty outcomes (e.g. precocious puberty, early thelarche, modification of cycle length, duration of menstrual bleeding, irregular menstrual periods, heavy menstrual flow, missed menstrual periods, spotting in the middle of a menstrual period, breast tenderness, frequency of pregnancies, and miscarriages or preterm deliveries) (Table 4; Fig. 13).

Fig. 13 Effect of soya infant formula on age of menarche. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

In 2010, a report about the possible association between uterine fibroids and SIF intake was published. In this cohort study with a low-to-moderate quality of evidence, the authors identified a risk ratio of 1·25, but with a CI of 0·97–1·61, associated with a non-significant P value( Reference D'Aloisio, Baird and DeRoo 162 ). With regard to SIF intake and potential association with endocrine dysfunction, interestingly, we found that most of these publications were published as case reports( Reference Brown, Peerson and Fontaine 43 , Reference Conrad, Chiu and Silverman 163 , Reference Mousavi, Tavakoli and Mardan 164 ). Messina et al. ( Reference Messina and Redmond 165 ) reported no association between SIF intake and thyroid function disturbances in healthy infants with euthyroidism. These investigators identified fourteen trials in which the effects of soya foods or isoflavones on at least one measure of thyroid function were evaluated in healthy subjects: eight included only women; four involved only men; two included both men and women. With only one exception, either no effects or only very limited changes were observed in these trials. Thus together, the findings provide little evidence that in euthyroid, iodine-replete subjects, soya foods or isoflavones adversely affect thyroid function( Reference Messina and Redmond 165 ).

Discussion

Soya has been used as food throughout the world for thousands of years. Ruhräh( Reference Ruhrah 1 ) published the first report on the use of a soyabean-based formula for infants in 1909. Early SIF contained soya flour, a constituent with a poorer protein digestibility and a reduced protein content when compared with the SPI used in modern SIF. SPI replaced soya flour in infant formulas during the early 1960s. In the 1970s, methionine, iodine, carnitine, taurine, choline and inositol were added to standard SIF. Modern SIF meet the AAP recommendations and the Infant Formula Act (1980 and subsequent amendments in 1986) requirements for term infants( Reference Ruhrah 1 4 ).

Approximately 25 % of infants in the USA are fed SIF at some point in their first year of life (AAP, 2008)( Reference Bhatia and Greer 155 ). Recently, some findings generated in animal models or human observations have challenged the use of these formulas in infants and children because of concerns about potential negative effects on growth, bone health, immunity, cognition, and reproductive or endocrine functions( Reference Merritt and Jenks 74 , Reference Vandenplas, De Greef and Devreker 106 ). The first review about soya was a narrative review published in 1988 that focused on growth and bone mineralisation. It was a result of concerns regarding adequate bone mineralisation when rickets was observed in very-low-birth-weight infants receiving soya-based feedings. This review concluded that children fed a soya isolate formula (old composition) had a pattern of growth similar to that of children fed a CMF and that infants fed a soya isolated formula had significantly lower bone mineral content and bone width at 3, 6, 9 and 12 months of age than those fed CMF, but that their values were similar to those of previously studied infants fed HM with vitamin D supplementation( 32 ). After the publication of this paper, at least eighteen additional narrative reviews on different aspects of safety and/or efficacy of SIF were published, most of them demonstrating a safety profile for use in children( 4 , Reference Chen and Rogan 9 , Reference Cantani and Lucenti 49 , Reference Zoppi and Guandalini 62 , Reference Mendez, Anthony and Arab 66 , Reference Miniello, Morol and Tarantino 69 , Reference Merritt and Jenks 74 , Reference Agostoni, Axelsson and Goulet 78 , Reference Turck 86 , Reference Song, Chun and Hwang 87 , Reference Johnson, Loomis and Flake 91 , Reference Badger, Gilchrist and Terry Pivik 100 , Reference Vandenplas, De Greef and Devreker 106 , Reference Dinsdale and Ward 117 , Reference Kattan, Cocco and Järvinen 121 , Reference Jefferson and Williams 125 , Reference Jefferson, Patisaul and Williams 128 , Reference Crinella 130 ). In addition to these publications, only three systematic reviews with a meta-analysis were published about the efficacy of soya as an adjuvant in acute diarrhoea, infantile colic or cows' milk protein allergy prevention. In these publications, Brown et al. ( Reference Brown, Peerson and Fontaine 43 ) assessed the effects of continued feeding of non-HM or formulas to infants during acute diarrhoea on their treatment failure rates, stool frequency and amount, diarrhoeal duration and body-weight change. They concluded that the vast majority of young children with acute diarrhoea can be successfully managed with continued feeding of undiluted non-HM. Lucassen et al. ( Reference Lucassen, Assendelft and Gubbels 55 ) concluded that in infants with infantile colic, the effectiveness of substitution with soya formula milks is unclear when only trials of good methodological quality are considered. Finally, Osborn & Sinn( Reference Ostrom, Jacobs and Merritt 82 ) concluded that soya formula feeding cannot be recommended for the prevention of allergy or food intolerance in infants at a high risk of allergy or food intolerance.

To our knowledge, this is the first systematic review with a meta-analysis published with the focus mainly on SIF and safety in infants and children. It has the advantage of covering evidence analysis from 1909 to July 2013 (104 years), including papers published on SIF, non-enriched SIF and supplemented/enriched SIF. This extensive analysis objectively showed that SIF intake in normal full-term infants – even during the most rapid phase of growth – is associated with normal anthropometric growth, adequate protein status, bone mineralisation and normal immune development. The importance of the meta-analysis reported herein is that data demonstrate the negative effects of the ‘old/unsupplemented soya formulas’ on Ca metabolism and bone mineral content. For example, Chan et al. ( Reference Chan, Leeper and Boo 138 ) studied the mineral metabolism in healthy term infants fed the old soya formula containing different sources of carbohydrates. Exclusively breast-fed infants served as controls. These investigators found that at 2 and 4 months of age, the breast-fed infants had higher bone mineral content and bone density. On the contrary, more recent studies using modern/supplemented SIF have shown growth patterns, Ca levels, bone mineral content, serum Hb levels, total protein levels, immune factors, and upper respiratory or diarrhoeic infection risk similar to those found with other types of feedings.

Few studies have evaluated the impact of SIF on neurodevelopment. For example, a study carried out by Malloy & Berendes( Reference Malloy and Berendes 148 ), in school-aged children who were fed either SIF or HM during their first year of life, showed no differences in intelligence quotient, behavioural problems, learning impairment or emotional problems. Strom et al. ( Reference Strom, Shinnar and Ziegler 149 ) conducted a study among adults aged 20–34 years who, as infants, participated in controlled feeding studies. Results indicated no differences in men or women with regard to the achievements of the level of college or trade school education, whether they were fed SIF or CMF. Andres et al. ( Reference Andres, Cleves and Bellando 150 ), in a more recent study in healthy infants, assessed the Bayley Scales of Infant Development and the Preschool Language Scale-3 during the first year of life. No differences were found between the CMF-fed and SIF-fed infants. We are aware of the debate about differences in behaviour (mental, psychomotor and language) in breast-fed infants and formula-fed infants, which are not necessarily related only to the type of feedings.

SPI contains 1–2 % of phytates, which may impair the absorption of minerals and trace elements. Modern SIF contain higher levels of micronutrients (Ca, Zn, Fe, etc.) when compared with CMF or HM. We found that feeding SIF to young infants did not result in any negative impact on the levels of Hb, Zn, Ca and overall growth (Figs. 1–3 5). Similarly, we also found that SIF contain significantly higher levels of aluminium than CMF and HM (SIF 500–2500 μg/l, CMF 15–400 μg/l and HM 4–65 μg/l). This systematic review did not find any evidence of a negative health effect of this metal in children. SIF should not be fed to preterm infants or infants with renal failure. Studies have concluded that in term infants with normal renal function, there is no risk of aluminium toxicity from SIF.

Finally, it is known that phyto-oestrogens represent a broad group of plant-derived compounds of non-steroidal structure that are abundant within the plant kingdom, including soya, and have a weak oestrogenic activity. Minimum data are available on the potential effects of exposure to phyto-oestrogens in young children on later sexual and reproductive development. SIF-fed infants may have higher serum and urine levels of genistein and daidzein. As has been mentioned earlier, it seems that most of the phyto-oestrogens present in the plasma of SIF-fed infants are in a conjugated form and are therefore unable to exert hormonal effects( Reference Hugget, Pridmore and Malnoe 158 ). The exhaustive analysis that we conducted in the present systematic review produced inconclusive results. We identified two cohort studies with a moderate quality of evidence of marginal adverse effects of SIF on early menarche. Furthermore, two other studies with a very low quality of evidence (one cross-sectional study and one case–control study) showed that SIF would be a risk factor for the presence of breast tissue during the second year of life. Additionally, one cohort study identified an association of SIF intake with a significant increase in the duration of women's menstrual cycles and more discomfort during menstrual periods. However, the same study did not show any statistical difference between the groups for more than thirty additional outcomes, such as presence of puberty, early thelarche, modification of cycle length, severity of menstrual flow, irregular menstrual periods, heavy menstrual flow, missed menstrual periods, spotting in the middle of a menstrual period, breast tenderness, frequency of pregnancies, and miscarriages or preterm deliveries.

This evidence analysis led us to establish that there is no significant effect of soya on important reproductive functions in human beings. The AAP has emphasised that literature reviews and clinical studies of infants fed SIF raise no clinical concerns with respect to nutritional adequacy, sexual development, thyroid disease, immune function or neurodevelopment. Additional studies confirm that SIF do not interfere with normal immune responses. The US Food and Drug Administration has also approved these formulas to be safe for use in infants.

Acknowledgements

The present study did not receive funding from any agency.

Y. V. is a consultant for United Pharmaceuticals and Biocodex. P. A. was a former employee of Abbott Nutrition (now retired). The other authors have no conflicts of interest to report.

References

1 Ruhrah, J (1909) The soy bean in infant feeding: preliminary report. Arch Pediatr 26, 496501.Google Scholar
2 Hill, LW & Stuart, HC (1928) A soy bean food preparation for feeding infants with milk idiosyncrasy. J Am Med Assoc 93, 985987.Google Scholar
3 Henley, EC & Kuster, JM (1994) Protein quality evaluation by protein digestibility-corrected amino acid scoring. Food Technol 48, 7477.Google Scholar
4 American Academy of Pediatrics (1998) Soy-protein formulas: recommendations for use in infant feeding. Pediatrics 101, 148153.CrossRefGoogle Scholar
5 Drugstore.com (2004) Formulation information for Isomil®, Isomil® Advance®, Isomil 2, Enfamil® ProSobee®, and Enfamil® Next Step® soy formulations. www.Drugstore.com. Google Scholar
6 USDA (2002) USDA-Iowa State University database on the isoflavone content of food, Release 1.3. http://www.nal.usda.gov/fnic/foodcomp/Data/isoflav/isoflav.html. United States Department of Agriculture and Iowa State University. Google Scholar
7 MAFF (1998) Plant Oestrogens in Soya-based Infant Formulae. http://archive.food.gov.uk/maff/archive/food/infsheet/1998/no167/167phy.htm. London: Ministry of Agriculture, Fisheries, and Food.Google Scholar
8 UK Committee on Toxicity (2003) Phytoestrogens and Health. http://www.food.gov.uk/multimedia/pdfs/phytoreport0503. London: Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment.Google Scholar
9 Chen, A & Rogan, WJ (2004) Isoflavones in soy infant formula: a review of evidence for endocrine and other activity in infants. Ann Rev Nutr 24, 3354.Google Scholar
10 Setchell, KD, Zimmer-Nechemias, L, Cai, J, et al. (1998) Isoflavone content of infant formulas and the metabolic fate of these phytoestrogens in early life. Am J Clin Nutr 68, 1453S1461S.Google Scholar
11 Tuohy, PG (2003) Soy infant formula and phytoestrogens. J Paediatr Child Health 39, 401405.Google Scholar
12 Essex, C (1996) Phytoestrogens and soy-based infant formula: risks remain theoretical. BMJ 313, 507508.Google Scholar
13 Wilczynski, NL, Haynes, RB & Hedges Team (2004) Developing optimal search strategies for detecting clinically sound prognostic studies in MEDLINE: an analytic survey. BMC Med 2, 23.Google Scholar
14 Atkins, D, Best, D, Briss, PA, et al. (2004) Grading quality of evidence and strength of recommendations. BMJ 328, 14901498.Google Scholar
15 Fomon, SJ (1959) Comparative study of human milk and a soya bean formula in promoting growth and nitrogen retention by infants. Pediatrics 24, 577584.Google Scholar
16 Shepard, TH, Pyne, GE, Kirschvink, JF, et al. (1960) Soybean goiter. N Engl J Med 262, 10991103.Google Scholar
17 Cowan, CC, Brownle, RC & deLoache, WR (1969) A soy protein isolate formula in the management of allergy in infants and children. South Med J 62, 389393.Google Scholar
18 Ament, ME & Rubin, CE (1972) Soy protein – another cause of the flat intestinal lesion. Gastroenterology 62, 227234.Google Scholar
19 Halpin, TC, Byrne, WJ & Ament, ME (1977) Colitis, persistent diarrhea and soy protein intolerance. J Pediatr 91, 404407.CrossRefGoogle ScholarPubMed
20 Powell, GK (1978) Milk- and soy-induced enterocolitis of infancy: clinical features and standardization of challenge. J Pediatr 93, 553560.Google Scholar
21 Naude, SP, Prinsloo, JG & Haupt, CE (1979) Comparison between a humanized cow's milk and a soy product for premature infants. S Afr Med J 55, 982986.Google Scholar
22 Zoppi, G, Zamboni, G, Bassani, N, et al. (1979) Gammaglobulin level and soy-protein intake in early infancy. Eur J Pediatr 131, 6169.Google Scholar
23 Shenai, JP, Jhaveri, BM, Reynolds, JW, et al. (1981) Nutritional balance studies in very-low-birth-weight infants: role of soy formula. Pediatrics 67, 631637.Google Scholar
24 Callenbach, JC, Sheehan, MB, Abramson, SJ, et al. (1981) Etiologic factors in rickets of very-low-birth-weight infants. J Pediatr 98, 800805.CrossRefGoogle ScholarPubMed
25 Gruskay, FL (1982) Comparison of breast, cow and soy feedings in the prevention of onset of allergic disease: a 15-year prospective study. Clin Pediatr (Phila) 21, 486491.Google Scholar
26 Poley, JR & Klein, AW (1983) Scanning electron microscopy of soy protein-induced damage of small bowel mucosa in infants. J Pediatr Gastroenterol Nutr 2, 271287.Google Scholar
27 Hall, RT, Callenbach, JC, Sheehan, MB, et al. (1984) Comparison of calcium- and phosphorus-supplemented soy isolate formula with whey-predominant premature formula in very-low-birth-weight infants. J Pediatr Gastroenterol Nutr 3, 571576.Google Scholar
28 Dagan, R, Gorodischer, R & Moses, SW (1984) Dietary treatment of acute diarrhea: comparison between cow's milk and a soy formula without disaccharides. J Trop Pediatr 30, 221224.Google Scholar
29 Kulkarni, PB, Dorand, RD, Bridger, WM, et al. (1984) Rickets in premature infants fed different formulas. South Med J 77, 1316.Google Scholar
30 Sutton, RE & Hamilton, JR (1968) Tolerance of young children with severe gastroenteritis to dietary lactose: a controlled study. Canad Med Assoc J 99, 980982.Google Scholar
31 Sampson, HA (1988) The role of food hypersensitivity and mediator release in atopic dermatitis. J Allergy Clin Immunol 81, 635645.Google Scholar
32 Nutrition Review Committee (1988) Bone mineralization and growth in term infants fed soy-based or cow milk-based formula. Nutr Rev 46, 152154.Google Scholar
33 Iyngkaran, N, Yadav, M, Looi, LM, et al. (1988) Effect of soy protein on the small bowel mucosa of young infants recovering from acute gastroenteritis. J Pediatr Gastroenterol Nutr 7, 6875.Google ScholarPubMed
34 Conway, SP & Ireson, AT (1989) Acute gastroenteritis in well-nourished infants: comparison of four feeding regimens. Arch Dis Child 64, 8791.Google Scholar
35 Chandra, RK, Singh, G & Shridhara, B (1989) Effect of feeding whey hydrolysate, soy and conventional cow milk formulas on incidence of atopic disease in high risk infants. Ann Allergy 63, 102106.Google Scholar
36 Cantani, A, Ferrara, M, Ragno, W, et al. (1990) Efficacy and safety of a soy-protein-formula for feeding babies with atopic dermatitis and cow milk hypersensitivity. Eur Rev Med Pharmacol Sci 12, 311318.Google Scholar
37 Bock, SA & Atkins, FM (1990) Patterns of food hypersensitivity during 16 years of double-blind, placebo-controlled food challenges. J Pediatr 117, 561567.Google Scholar
38 Willoughby, A, Graubard, BI, Hocker, A, et al. (1990) Population-based study of the developmental outcome of children exposed to chloride-deficient infant formula. Pediatrics 85, 485490.Google Scholar
39 Malloy, MH, Willoughby, A, Graubard, B, et al. (1990) Exposure to a chloride-deficient formula during infancy: outcome at ages 9 and 10 years. Pediatrics 86, 601610.Google Scholar
40 Giampietro, PG, Ragno, V, Daniele, S, et al. (1992) Soy hypersensitivity in children with food allergy. Ann Allergy 69, 143146.Google ScholarPubMed
41 Buts, JP, Di Sano, C & Hansdorffer, S (1993) Clinical evaluation of the tolerance for a soy-based special milk formula in children with cow's milk protein intolerance/allergy (CMPI/CMPA). Minerva Pediatr 45, 209213.Google Scholar
42 Churella, HR, Borschel, MW, Thomas, MR, et al. (1994) Growth and protein status of term infants fed soy protein formulas differing in protein content. J Am Coll Nutr 13, 262267.Google Scholar
43 Brown, KH, Peerson, JM & Fontaine, O (1994) Use of nonhuman milks in the dietary management of young children with acute diarrhea meta-analysis of clinical trials. Pediatrics 93, 1727.Google Scholar
44 Burks, AW, Castee, HB, Fiedorek, SC, et al. (1994) Prospective oral food challenge study of two soybean protein isolates in patients with possible milk or soy protein enterocolitis. Pediatr Allergy Immunol 5, 4045.Google Scholar
45 Chorazy, PA, Himelhoch, S, Hopwood, NJ, et al. (1995) Persistent hypothyroidism in an infant receiving a soy formula: case report and review of the literature. Pediatrics 1, 148150.Google Scholar
46 Magnolfi, C, Zani, G, Lacava, L, et al. (1996) Soy allergy in atopic children. Ann Allergy Asthma Immunol 77, 197201.Google Scholar
47 Bruno, G, Giampietro, PG, Del Guercio, MJ, et al. (1997) Soy allergy is not common in atopic children: a multicenter study. Pediatr Allergy Immunol 8, 190193.Google Scholar
48 Jabbar, MA, Larrea, J & Shaw, RA (1997) Abnormal thyroid function tests in infants with congenital hypothyroidism: the influence of soy-based formula. J Am Coll Nutr 16, 280282.CrossRefGoogle ScholarPubMed
49 Cantani, A & Lucenti, P (1997) Natural history of soy allergy and/or intolerance in children, and clinical use of soy protein formulas. Pediatr Allergy Immunol 8, 5974.Google Scholar
50 Vanderhoof, JA, Murray, ND, Paule, CL, et al. (1997) Use of soy fiber in acute diarrhea in infants and toddlers. Clin Pediatr (Phila) 36, 135139.Google Scholar
51 Kuiper, JM, Lemmen, JG, Carlsson, B, et al. (1998) Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor. Endocrinology 139, 42524263.Google Scholar
52 Businco, L, Bruno, G & Giampietro, PG (1998) Soy protein for the prevention and treatment of children with cow-milk allergy. Am J Clin Nutr 68, Suppl. 6, 1447S1452S.Google Scholar
53 Quak, SH & Tan, SP (1998) Use of soy-protein formulas and soyfood for feeding infants and children in Asia. Am J Clin Nutr 68, Suppl. 6, 1444S1446S.Google Scholar
54 Irvine, CHG, Shand, N, Fitzpatrick, MG, et al. (1998) Daily intake and urinary excretion of genistein and daidzein by infants fed soy- or dairy-based infant formulas. Am J Clin Nutr 68, Suppl. 6, 1462S1465S.Google Scholar
55 Lucassen, PLBJ, Assendelft, WJJ, Gubbels, JW, et al. (1998) Effectiveness of treatments for infantile colic: systematic review. BMJ 316, 15631569.Google Scholar
56 Burks, WA, James, JM, Hiegel, A, et al. (1998) Atopic dermatitis and food hypersensitivity reactions. J Pediatr 132, 132136.CrossRefGoogle ScholarPubMed
57 Sheehan, DM (1998) Herbal medicines, phytoestrogens and toxicity: risk:benefit considerations. Proc Soc Exp Biol Med 217, 379385.Google Scholar
58 Irvine, CH, Fitzpatrick, MG & Alexander, SL (1998) Phytoestrogens in soy-based infant foods: concentrations, daily intake, and possible biological effects. Proc Soc Exp Biol Med 3, 247253.Google Scholar
59 Fayad, IM, Hashem, M, Hussein, A, et al. (1999) Comparison of soy-based formulas with lactose and with sucrose in the treatment of acute diarrhea in infants. Arch Pediatr Adolesc Med 153, 675680.Google Scholar
60 Zeiger, RS, Sampson, HA, Bock, S, et al. (1999) Soy allergy in infants and children with IgE-associated cow's milk allergy. J Pediatr 134, 614622.Google Scholar
61 Badger, TM, Ronis, MJJ, Hakkak, R, et al. (2002) The health consequences of early soy consumption. J Nutr 132, 559S565S.Google Scholar
62 Zoppi, G & Guandalini, S (1999) The story of soy formula feeding in infants: a road paved with good intentions. J Pediatr Gastroenterol Nutr 28, 541543.Google Scholar
63 Setchell, KDR (2000) Absorption and metabolism of soy isoflavones – from food to dietary supplements and adults to infants. J Nutr 130, 654S655S.Google Scholar
64 Goldman, LR, Newbold, R & Swan, SH (2001) Exposure to soy-based formula in infancy. JAMA 286, 24022403.Google Scholar
65 Barrett, JR (2002) Soy and children's health: a formula for trouble. Environ Health Perspect 10, A294A296.Google Scholar
66 Mendez, MA, Anthony, MS & Arab, L (2002) Soy-based formulae and infant growth and development: a review. J Nutr 132, 21272130.Google Scholar
67 Ostrom, K, Borschel, MW, Westcott, JE, et al. (2002) Lower calcium absorption in infants fed casein hydrolysate- and soy protein-based infant formulas containing palm olein versus formulas without palm olein. J Am Coll Nutr 21, 564569.Google Scholar
68 Klemola, T, Vanto, T, Juntunen-Backman, K, et al. (2002) Allergy to soy formula and to extensively hydrolyzed whey formula in infants with cow's milk allergy: a prospective, randomized study with a follow-up to the age of 2 years. J Pediatr 140, 219224.Google Scholar
69 Miniello, VL, Morol, GE, Tarantino, M, et al. (2003) Soy-based formulas and phyto-oestrogens: a safety profile. Acta Paediatr Scand 441, 93100.Google Scholar
70 Ahn, KM, Han, YS, Nam, SY, et al. (2003) Prevalence of soy protein hypersensitivity in cow's milk protein-sensitive children in Korea. J Korean Med Sci 18, 473477.Google Scholar
71 Stettler, N, Stallings, VA, Troxel, AB, et al. (2005) Weight gain in the first week of life and overweight in adulthood: a cohort study of European American subjects fed infant formula. Circulation 111, 18971903.Google Scholar
72 Hoey, L, Rowland, IR, Lloyd, AS, et al. (2004) Influence of soya-based infant formula consumption on isoflavone and gut microflora metabolite concentrations in urine and on faecal microflora composition and metabolic activity in infants and children. Br J Nutr 91, 607616.Google Scholar
73 Giampietro, PG, Bruno, G, Furcolo, G, et al. (2004) Soy protein formulas in children: no hormonal effects in long-term feeding. J Pediatr Endocrinol Metab 17, 191196.Google Scholar
74 Merritt, RJ & Jenks, BH (2004) Safety of soy-based infant formulas containing isoflavones: the clinical evidence. J Nutr 134, 1220S1224S.Google Scholar
75 Hays, T & Wood, RA (2005) A systematic review of the role of hydrolyzed infant formulas in allergy prevention. Arch Pediatr Adolesc Med 159, 810816.Google Scholar
76 Berger-Achituv, S, Shohat, T, Romano-Zelekha, O, et al. (2005) Widespread use of soy-based formula without clinical indications. J Ped Gastroenterol Nutr 41, 660666.Google Scholar
77 Klemola, T, Kalimo, K, Poussa, T, et al. (2005) Feeding a soy formula to children with cow's milk allergy: the development of immunoglobulin E-mediated allergy to soy and peanuts. J Pediatr Allergy Immunol 16, 641646.Google Scholar
78 Agostoni, C, Axelsson, I, Goulet, O, et al. (2006) Soy protein infant formulae and follow-on formulae: a commentary by the ESPGHAN Committee on Nutrition. J Ped Gastroenterol Nutr 42, 352361.Google Scholar
79 Pedrosa, M, Pascual, CY, Larco, JL, et al. (2006) Palatability of hydrolysates and other substitution formulas for cow's milk-allergic children: a comparative study of taste, smell, and texture evaluated by healthy volunteer. J Investig Allergol Clin Immunol 16, 351356.Google Scholar
80 D'Auria, E (2006) Impact of soy formulas on growth. J Ped Gastroenterol Nutr 42, 594595.Google Scholar
81 Osborn, DA & Sinn, J (2006) Soy formula for prevention of allergy and food intolerance in infants. Cochrane Database Syst Rev, issue 4, CD003741.Google Scholar
82 Ostrom, K, Jacobs, JR, Merritt, RJ, et al. (2006) Decreased regurgitation with a soy formula containing added soy fiber. Clin Pediatr (Phila) 45, 2936.Google Scholar
83 Ballmer-Weber, B, Holzhauser, T, Scibilia, J, et al. (2007) Clinical characteristics of soybean allergy in Europe: a double-blind, placebo-controlled food challenge study. J Allergy Clin Immunol 119, 14891496.Google Scholar
84 Fortes, E, Malerba, M, Luchini, P, et al. (2007) Ingestão excessiva de fitoestrógenos e telarca precoce: relato de caso compossível correlação (Excessive ingestion of phyto-oestrogens and precocious thelarche: case report with a possible correlation). Arq Bras Endocrinol Metab 51, 500503.Google Scholar
85 Halm, BM, Ashburn, LA & Franke, AA (2007) Isoflavones from soya foods are more bioavailable in children than adults. Br J Nutr 98, 9981005.Google Scholar
86 Turck, D (2007) Soy protein for infant feeding: what do we know? Curr Opin Clin Nutr Metab Care 10, 360365.Google Scholar
87 Song, WO, Chun, OK, Hwang, I, et al. (2007) Soy isoflavones as safe functional ingredients. J Med Food 10, 571580.Google Scholar
88 Agostoni, C, Fiocchi, A, Riva, E, et al. (2007) Growth of infants with IgE-mediated cow's milk allergy fed different formulas in the complementary feeding period. Pediatr Allergy Immunol 18, 599606.Google Scholar
89 Wolff, MS, Britton, JA, Boguski, L, et al. (2008) Environmental exposures and puberty in inner-city girls. Environ Res 107, 393400.Google Scholar
90 Zuidmeer, L, Goldhahn, K, Rona, R, et al. (2008) The prevalence of plant food allergies: a systematic review. J All Clin Immunol 121, 12101218.Google Scholar
91 Johnson, K, Loomis, G, Flake, D, et al. (2008) Effects of soy protein-based formula in full-term infants. Am Fam Physician 77, 8788.Google ScholarPubMed
92 Ngamphaiboon, J, Chatchatee, P & Thongkaew, T (2008) Cow's milk allergy in Thai children. Asian Pacif J Allerg Immunol 26, 199204.Google Scholar
93 Mehr, S & Kemp, A (2008) Feeding choice for children with immediate allergic reactions to cow's milk protein. Med J Austr 189, 178179.Google Scholar
94 Boucher, BA, Cotterchio, M, Kreiger, N, et al. (2008) Soy formula and breast cancer risk. Epidemiology 19, 165166.Google Scholar
95 Kemp, A, Hill, D, Allen, K, et al. (2008) Guidelines for the use of infant formulas to treat cows milk protein allergy: an Australian consensus panel opinion. Med J Austr 188, 109112.Google Scholar
96 Bernbaum, J, Umbach, D, Ragan, NB, et al. (2008) Pilot studies of estrogen-related physical findings in infants. Environ Health Perspect 116, 416420.Google Scholar
97 Koplin, J, Dharmage, S, Gurrin, L, et al. (2008) Soy consumption is not a risk factor for peanut sensitization. J Allerg Clin Immunol 121, 14551459.Google Scholar
98 Caminiti, L, Passalacqua, LG, Barberi, S, et al. (2009) A new protocol for specific oral tolerance induction in children with IgE-mediated cow's milk allergy. Asthma Allergy Proc 30, 443448.Google Scholar
99 Antunes, J, Borrego, LM, Queiroz, A, et al. (2009) Allergy to extensively hydrolysed formulas. Allergol Immunopathol (Madr) 37, 272278.Google Scholar
100 Badger, T, Gilchrist, J, Terry Pivik, R, et al. (2009) The health implications of soy infant formula. Am J Clin Nutr 89, 1668S1672S.Google Scholar
101 Lee, SA, Shu, XO, Li, H, et al. (2009) Adolescent and adult soy food intake and breast cancer risk: results from the Shanghai Women's Health Study. Am J Clin Nutr 89, 19201926.Google Scholar
102 Korde, L, Wu, AH & Fears, T (2009) Childhood soy intake and breast cancer risk in Asian American women. Cancer Epidemiol Biomarkers Prev 18, 10501059.Google Scholar
103 Guest, JF & Nagy, E (2009) Modelling the resource implications and budget impact of managing cow milk allergy in Australia. Curr Med Res Opin 25, 339349.Google Scholar
104 Palmer, J, Herbst, A, Noller, K, et al. (2009) Urogenital abnormalities in men exposed to diethylstilbestrol in utero: a cohort study. Environ Health 8, 37.Google Scholar
105 Cederroth, CH, Zimmermann, A, Eustache, F, et al. (2010) Soy, phyto-oestrogens and male reproductive function: a review. Int J Andrology 33, 304316.Google Scholar
106 Vandenplas, Y, De Greef, E, Devreker, T, et al. (2011) Soy infant formula: is it that bad? Acta Pædiatr 100, 162166.Google Scholar
107 Dias, A, Santos, A & Pinheiro, JA (2010) Persistence of cow's milk allergy beyond two years of age. Allergol Immunopathol (Madr) 38, 812.Google Scholar
108 Bolca, S, Urpi-Sarda, M, Blondeel, P, et al. (2010) Disposition of soy isoflavones in normal human breast tissue. Am J Clin Nutr 91, 976984.Google Scholar
109 Cheng, G, Remer, T, Prinz-Langenohl, R, et al. (2010) Relation of isoflavones and fiber intake in childhood to the timing of puberty. Am J Clin Nutr 92, 556564.Google Scholar
110 Terracciano, L, Bouygue, GR, Sarratud, T, et al. (2010) Impact of dietary regimen on the duration of cow's milk allergy: a random allocation study. Clin Exp Allergy 40, 637642.Google Scholar
111 Tillet, T (2010) Soy formula of “minimal concern”. Environ Health Perspect 118, A335A336.Google Scholar
112 Nachmias, M, Landman, Y, Danon, Y, et al. (2010) Soy allergy following early soy feeding in neonates. Isr Med Assoc J 12, 684686.Google Scholar
113 Sladkevicius, E, Nagy, E, Lack, G, et al. (2010) Resource implications and budget impact of managing cow milk allergy in the UK. J Med Econ 13, 119128.Google Scholar
114 Katz, Y, Rajuan, N, Goldberg, M, et al. (2010) Early exposure to cow's milk protein is protective against IgE-mediated cow's milk protein allergy. J Allergy Clin Immunol 126, 7782.Google Scholar
115 Patisaul, H & Jefferson, W (2010) The pros and cons of phytoestrogens. Front Neuroendocrinol 31, 400419.Google Scholar
116 Donovan, S, Andres, A, Mathai, RA, et al. (2010) Soy formula and isoflavones and the developing intestine. Nutr Rev 67, S192S200.Google Scholar
117 Dinsdale, E & Ward, W (2010) Early exposure to soy isoflavones and effects on reproductive health: a review of human and animal studies. Nutrients 2, 11561187.Google Scholar
118 Wada, K, Nakamura, K, Masue, T, et al. (2011) Soy intake and urinary sex hormone levels in preschool Japanese children. Am J Epidemiol 178, 9981003.Google Scholar
119 McCarver, G, Bhatia, J, Chambers, C, et al. (2011) NTP-CERHR expert panel report on the developmental toxicity of soy infant formula. Birth Defects Res B Dev Reprod Toxicol 92, 421468.Google Scholar
120 Kim, J, Kim, S, Huh, K, et al. (2011) High serum isoflavone concentrations are associated with the risk of precocious puberty in Korean girls. Clin Endocr (Oxf) 75, 831835.Google Scholar
121 Kattan, JD, Cocco, RR & Järvinen, KM (2011) Milk and soy allergy. Pediatr Clin N Am 58, 407426.Google Scholar
122 Dabeka, R, Fouquet, A, Belisle, S, et al. (2011) Lead, cadmium and aluminum in Canadian infant formulae, oral electrolytes and glucose solutions. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 28, 744753.Google Scholar
123 Degen, G, Blaszkewicz, M, Shi, L, et al. (2011) Urinary isoflavone phytoestrogens in German children and adolescents – a longitudinal examination in the DONALD cohort. Mol Nutr Food Res 55, 359367.Google Scholar
124 Nguyen, R, Umbach, D, Parad, R, et al. (2011) US assessment of estrogen-responsive organ growth among healthy term infants: piloting methods for assessing estrogenic activity. Pediatr Radiol 41, 633642.Google Scholar
125 Jefferson, W & Williams, C (2011) Circulating levels of genistein in the neonate, apart from dose and route, predict future adverse female reproductive outcomes. Reprod Toxicol 31, 272279.Google Scholar
126 Durham, LE (2011) Food allergies in children. Don't forget allergy in eczema. BMJ 8, 342.Google Scholar
127 Levy, SA, Dortas Junior, SD, Pires, AH, et al. (2012) Atopy patch test (APT) in the diagnosis of food allergy in children with atopic dermatitis. An Bras Dermatol 87, 724728.Google Scholar
128 Jefferson, W, Patisaul, H & Williams, C (2012) Reproductive consequences of developmental phytoestrogen exposure. Reproduction 143, 247260.Google Scholar
129 Blom, WM, Vlieg-Boerstra, B, Kruizinga, A, et al. (2013) Threshold dose distributions for 5 major allergenic foods in children. J Allergy Clin Immunol 131, 172179.Google Scholar
130 Crinella, F (2012) Does soy-based infant formula cause ADHD? Update and public policy considerations. Expert Rev Neurother 12, 395407.Google Scholar
131 Kay, JL, Daeschner, CW Jr & Desmond, MM (1960) Evaluation of infants fed soybean and evaporated milk formulae from birth to three months. A comparison of weight, length, hemoglobin, hematocrit, plasma biochemical values. Am J Dis Child 100, 264276.Google Scholar
132 Cherry, FF, Cooper, MD, Stewart, RA, et al. (1968) Cow versus soy formulas. Comparative evaluation in normal infants. Am J Dis Child 115, 677692.Google Scholar
133 Sellars, WA, Halpern, SR, Johnson, RB, et al. (1971) New growth charts: soy, cow and breast milk comparison. Ann Allergy 29, 126134.Google Scholar
134 Dean, ME (1973) Study of normal infants fed a soya protein isolate formula. Med J Aust 1, 12891293.Google Scholar
135 Jung, AL & Carr, SL (1977) A soy protein formula and a milk-based formula. A comparative evaluation in milk-tolerant infants showed no significant nutritional differences. Clin Pediatr (Phila) 16, 982985.Google Scholar
136 Zoppi, G, Gerosa, F, Pezzini, A, et al. (1982) Immunocompetence and dietary protein intake in early infancy. J Pediatr Gastroenterol Nutr 1, 175182.Google Scholar
137 Khöler, L, Meeuwisse, G & Mortensson, W (1984) Food intake and growth of infants between six and twenty-six weeks of age on breast milk, cow's milk formula or soy formula. Acta Paediatr Scand 73, 40484052.Google Scholar
138 Chan, GM, Leeper, L & Boo, LS (1987) Effects of soy formulas on mineral metabolism in term infants. Am J Dis Child 141, 527530.Google Scholar
139 Steichen, JJ & Tsang, RC (1987) Bone mineralization and growth in term infants fed soy-based or cow milk-based formula. J Pediatr 110, 687692.Google Scholar
140 Hillman, LS, Chow, W, Salmons, SS, et al. (1988) Vitamin D metabolism, mineral homeostasis and bone mineralization in term infants fed human milk, cow milk-based formula or soy-based formula. J Pediatr 112, 864874.Google Scholar
141 Venkataraman, PS, Luhar, H & Neylan, MJ (1992) Bone mineral metabolism in full-term infants fed human milk, cow milk-based and soy-based formulas. Am J Dis Child 146, 13021305.Google Scholar
142 Mimouni, F, Campaigne, B, Neylan, M, et al. (1993) Bone mineralization in the first year of life in infants fed human milk, cow-milk formula or soy-based formula. J Pediatr 122, 348354.Google Scholar
143 Giovannini, M, Agostoni, C, Fiocchi, A, et al. (1994) Antigen-reduced infant formulas versus human milk: growth and metabolic parameters in the first six months of life. J Am Coll Nutr 13, 357363.Google Scholar
144 Lasekan, JB, Ostrom, KM, Jacobs, JR, et al. (1999) Growth of newborn, term infants fed soy formulas for one year. Clin Pediatr 38, 563571.Google Scholar
145 Seppo, L, Korpela, R, Lönnerdal, B, et al. (2005) A follow-up study of nutrient intake, nutritional status and growth in infants with cow milk allergy fed either a soy formula or an extensively hydrolyzed whey formula. Am J Clin Nutr 82, 140145.Google Scholar
146 Han, YH, Yon, M, Han, HS, et al. (2011) Zinc status and growth of Korean infants fed human milk, casein-based or soy-based formula: three-year longitudinal study. Nutr Res Pract 5, 4651.Google Scholar
147 Andres, A, Casey, PH, Cleves, MA, et al. (2013) Body fat and bone mineral content of infants fed breast milk, cow's milk formula or soy formula during the first year of life. J Pediatr 163, 4954.Google Scholar
148 Malloy, MH & Berendes, H (1998) Does breastfeeding influence intelligence quotients at 9 and 10 years of age? Early Hum Dev 50, 209217.Google Scholar
149 Strom, BL, Shinnar, R, Ziegler, EE, et al. (2001) Exposure to soy-based formula in infancy and endocrinological and reproductive outcomes in young adulthood. JAMA 286, 807814.Google Scholar
150 Andres, A, Cleves, MA, Bellando, JB, et al. (2012) Developmental status of one-year-old infants fed breast milk, cow's milk formula or soy formula. Pediatrics 129, 11341140.Google Scholar
151 Zoppi, G, Gasparini, R, Mantovanelli, F, et al. (1983) Diet and antibody response to vaccinations in healthy infants. Lancet ii, 1114.Google Scholar
152 Businco, L, Bruno, G, Grandolfo, ME, et al. (1990) Response to poliovirus immunization and type of feeding in babies of atopic families. Pediatr Allergy Immunol 1, 6063.Google Scholar
153 Ostrom, KM, Cordle, CT, Schaller, JP, et al. (2002) Immune status of infants fed soy-based formulas with or without added nucleotides for 1 year: part 1: vaccine responses and morbidity. J Pediatr Gastroenterol Nutr 34, 137144.Google Scholar
154 Cordle, CT, Winship, TR, Schaller, JP, et al. (2002) Immune status of infants fed soy-based formulas with or without added nucleotides for one year: part 2: immune cell populations. J Pediatr Gastroenterol Nutr 34, 145153.Google Scholar
155 Bhatia, J & Greer, F (2008) Use of soy protein-based formulas in infant feeding. Pediatrics 121, 10621068.Google Scholar
156 Setchell, KD, Zimmer-Nechemias, L, Cai, J, et al. (1997) Exposure of infants to phyto-oestrogens from soy-based infant formula. Lancet 350, 2327.Google Scholar
157 Cao, Y, Calafat, AM, Doerge, DR, et al. (2009) Isoflavones in urine, saliva and blood of infants: data from a pilot study on the estrogenic activity of soy formula. J Expo Sci Environ Epidemiol 19, 223234.Google Scholar
158 Hugget, AC, Pridmore, S, Malnoe, A, et al. (1997) Phyto-oestrogens in soy-based infant formula. Lancet 350, 815816.Google Scholar
159 Adgent, MA, Daniels, JL, Rogan, WJ, et al. (2012) Early-life soy exposure and age at menarche. Paediatr Perinat Epidemiol 26, 163175.Google Scholar
160 Zung, A, Glaser, T, Kerem, Z, et al. (2008) Breast development in the first two years of life: an association with soy-based infant formulas. J Pediatr Gastroenterol Nutr 46, 191195.Google Scholar
161 Lambertina, W, Freni-Titulaer, MSPH, Cordero, J, et al. (1986) Premature thelarche in Puerto Rico. Am J Dis Child 140, 12631267.Google Scholar
162 D'Aloisio, AA, Baird, DD, DeRoo, LA, et al. (2010) Association of intrauterine and early-life exposures with diagnosis of uterine leiomyomata by 35 years of age in the sister study. Environ Health Perspect 118, 375381.Google Scholar
163 Conrad, SC, Chiu, H & Silverman, BL (2004) Soy formula complicates management of congenital hypothyroidism. Arch Dis Child 89, 3740.Google Scholar
164 Mousavi, SM, Tavakoli, N & Mardan, F (2006) Risk factors for goiter in primary school girls in Qom city of Iran. Eur J Clin Nutr 60, 426433.Google Scholar
165 Messina, M & Redmond, G (2006) Effects of soy protein and soybean isoflavones on thyroid function in healthy adults and hypothyroid patients: a review of the relevant literature. Thyroid 16, 249258.Google Scholar
Figure 0

Table 1 Studies excluded from the review

Figure 1

Table 2 Evidence from studies included in the review (weight, length, bone health and other nutritional parameters) (Standardised mean difference (SMD) values and 95 % confidence intervals)

Figure 2

Fig. 1 Effect of soya infant formula on weight gain. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 3

Fig. 2 Effect of soya infant formula on height gain. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 4

Fig. 3 Effect of soya infant formula on Hb values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 5

Fig. 4 Effect of soya infant formula on serum total proteins. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 6

Fig. 5 Effect of soya infant formula on serum zinc values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 7

Fig. 6 Effect of soya infant formula on total calcium values. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 8

Fig. 7 Effect of soya infant formula on bone mineral content (gm/cm2). SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 9

Table 3 Evidence from studies included in the review (immunity and infection risk) (Standardised mean difference (SMD) values and 95 % confidence intervals)

Figure 10

Fig. 8 Effect of soya infant formula on polio antibodies. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 11

Fig. 9 Effect of soya infant formula on diphtheria antibodies. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 12

Fig. 10 Effect of soya infant formula on infectious episodes/child. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 13

Table 4 Evidence from studies included in the review (reproductive and endocrine functions). (Odds ratios, risk ratios (RR) or standardised mean difference (SMD), weighted mean difference (WMD) values and 95 % confidence intervals)

Figure 14

Fig. 11 Effect of soya infant formula on genistein levels in serum. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 15

Fig. 12 Effect of soya infant formula on daidzein levels in serum. SMD, standardised mean difference; RCTSB, randomised controlled trial, single blind. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).

Figure 16

Fig. 13 Effect of soya infant formula on age of menarche. SMD, standardised mean difference. (A colour version of this figure can be found online at http://www.journals.cambridge.org/bjn).