Ultrasonographic Patterns of Reproductive Organs in Fed Soy Formula: Comparisons to Infants Fed and Milk Formula

Janet M. Gilchrist, PhD, RD, Mary Beth Moore, MD, MS, Aline Andres, PhD, Judy A. Estroff, MD, and Thomas M. Badger, PhD

Objective To determine if differences exist in -sensitive organ size between infants who were fed soy formula (SF), milk formula (MF), or (BF). Study design Breast buds, uterus, , prostate, and testicular volumes were assessed by ultrasonography in 40 BF, 41 MF, and 39 SF infants at age 4 months. Results There were no significant feeding group effects in anthropometric or body composition. Among girls, there were no feeding group differences in breast bud or uterine volume. MF infants had greater (P < .05) mean ovar- ian volume and greater (P < .01) numbers of ovarian cysts per ovary than did BF infants. Among boys, there were no feeding group differences in prostate or breast bud volumes. Mean testicular volume did not differ between SF and MF boys, but both formula-fed groups had lower volumes than BF infants. Conclusions Our data do not support major diet-related differences in reproductive organ size as measured by ultrasound in infants at age 4 months, although there is some evidence that ovarian development may be advanced in MF-fed infants and that testicular development may be slower in both MF and SF infants as compared with BF. There was no evidence that feeding SF exerts any estrogenic effects on reproductive organs studied. (J Pediatr 2010;156:215-20).

See related articles, p 209 and p 221

stimates suggest that soy-based formulas (SF) are fed to 20% to 25% of all formula-fed infants in the United States at some point during the first year of life,1,2 amounting to more than 1 million infants each year. Concerns have been Eraised about potential adverse consequences of SF related to estrogenic effects of isoflavones associated with the soy pro- tein isolate used as their sole protein source.2 Similar reports have appeared in the lay press and on Internet web sites. Although the American Academy of (AAP) recommends breast-feeding (BF) rather than formula feeding and cow’s milk formula (MF) over SF, it does not find any adverse effects of SF.3 The AAP position is essentially that milk formula, although not as good as human breast milk, has been in use for decades, has been reformulated several times to meet nutrient and safety requirements, and has a long track record of safety and efficacy. Therefore, the AAP does not see an advantage of SF over milk formula for general use and recommends SF only for a few medical indications (infants with galactosemia, hereditary lactase deficiency, and secondary ) as well as in situations in which a vegetarian diet is preferred. This stance is prudent in view of international controversy over SF, and it provides the middle ground of ‘‘error on the side of safety without appropriate data in children.’’ This is also the standpoint taken by Australia, Canada, , Ireland, New Zealand, Switzer- land, and the United Kingdom, where SF use is limited because of concerns about potential estrogenic effects of soy isofla- vones.3,4 One clear message from the AAP is that breast feeding is the gold standard against which MF is compared, and MF is the standard against which SF is compared. A recent report of an expert panel selected by the National Toxicology Program Center for the Evaluation of Risks to (NTP-CERHR) concluded that ‘‘There are insufficient human or experimental animal data available to permit a determination of the developmental or reproductive toxicity of soy .’’3 Several studies in rodents have shown that purified isoflavones added to the diet will affect development of reproductive tissues, including the and uterus.5-9 However, very few studies have evaluated the potential effects of SF on human reproduction, and data from those reports are difficult to interpret or ex- 10-12 From the Arkansas Children’s Nutrition Center (J.M.G., trapolate to human infants. Two recent studies in human infants that as- A.A., T.M.B.), Departments of Pediatrics (A.A., T.M.B.) and Radiology (M.B.M.), University of Arkansas for sessed breast development were based on physical examination by palpation Medical Sciences, Little Rock, AR; and Department of with comparison to a standardized object (beads/coin) for assessing breast Radiology (J.A.E.), Children’s Hospital Boston, Harvard 11,12 Medical School, Boston, MA size. Both studies suggest SF effects on breast tissue, but there were Supported by USDA CRIS # 6251-51000-005-03S. multiple limitations, including examiners were not blinded to infant feeding T.M.B. was a scientific advisor for the Soy Nutrition Institute. The other authors declare no conflicts of interest. Participants were in The Beginnings Study, which is BF Breast-fed registered as a clinical trial at www.clinicaltrial.gov.ID#: NCT00616395. MF Milk-based formula–fed SF Soy-based formula–fed 0022-3476/$ - see front matter. Copyright Ó 2010 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2009.08.043

215 THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 156, No. 2 group, sample sizes were small, and the methods used were study of infant feeding.13,14 Sonograms were obtained using too crude to accurately and objectively assess breast develop- an ACUSON Sequoia unit (Siemens Medical Solutions ment. Thus, there is a dearth of comparative data on repro- USA, Inc., Malvern, Pennsylvania) and the following trans- ductive tissues in children fed the 3 most widely used ducers: high-frequency 8.0-MHz linear array for breast methods of infant feeding (BF, MF, and SF). buds; 8.0-MHz vector for prostate, ovaries, and uterus; 15- The purpose of the current study was to use ultrasonog- MHz high-frequency linear array for . Images were raphy to assess the size of breast buds, uterus, ovaries, pros- stored in the hospital’s PAACS system (DR Systems, Inc., tate, and testes in well-characterized human infants San Diego, California) and were reviewed and measured by exclusively fed the 3 most common infant diets—BF, MF, one Board Certified Radiologist (M.B.M.). The sonographers and SF. and the radiologist were blinded to participant treatment group. All organs were measured in triplicate and in 3 or- Methods thogonal planes: sagittal (SAG), transverse (TRV), and ante- rioposterior (AP). For paired organs (ie, breast buds, ovaries, Participants for this study were 4 months old and were drawn and testes), measures were obtained in triplicate on each side. from an ongoing, longitudinal cohort, the Beginnings Study An average of the 3 values was calculated for each of the (www.clinicaltrials.gov, ID # NCT00616395). had 3 views.13,14 a healthy , with no diagnoses (eg, diabetes, pre- Uterus length (SAG) was measured to include the fundus eclampsia) or medications (eg, SSRIs, thyroid replacement) and cervix; the AP diameter was measured within the fundus thought to affect infant growth or development. All mothers or midportion. Uterine characterizations were ‘‘pear- were non-smokers and denied alcohol use during pregnancy. shaped’’ (fundus smaller than cervix), ‘‘cylindrical’’ (fundus In addition, mothers reported no use of any soy products or equal to cervix), or ‘‘heart-shaped’’ (fundus larger than cer- other estrogenic compounds during pregnancy and/or lacta- vix). The presence or absence of a visible endometrial stripe tion. Infants were term (>38 weeks), had an appropriate-for- was noted. During measurement of the ovaries, all identified weight (6 to 9 pounds), and had no follicles/cysts were measured. Macrocysts were defined as medical diagnoses or medications thought to affect growth cysts >1.0 cm. During measurement of the testes, the position or development. chose the feeding method for their of each was noted as inguinal or scrotal. , that is, BF, MF, or SF. All BF infants were fed breast Breast volume was calculated as SAG * TRV * AP. Volumes milk exclusively for 4 months. Fifty-four percent of the MF of prostate, uterus, and ovaries were calculated by using the infants were exclusively fed MF from birth, 41% switched formula for a prolate ellipse: 0.523 * SAG * TRV * AP. Testic- from BF to MF within 4 weeks, and 5% switched between 4 ular volume was calculated using the empiric formula of and 8 weeks. Of the SF infants, 23% were exclusively fed SF Lambert15: SAG * TRV * AP * 0.71. Ovarian and testicular from birth, 45% were switched to exclusive SF feeding within volumes were calculated as the mean values for the right 4 weeks, and 32% switched between 4 and 8 weeks. Parents and left organs. also chose the formula and used Enfamil Prosobee Lipil All values in the Tables and Figure are given as mean (soy formula) (Mead Johnson & Johnson and Co., Evansville, Æ SEM unless otherwise stated. Ultrasonographic data are Indiana), Enfamil Advance (milk formula) (Mead Johnson & plotted against body weight, and the Student t test was Johnson and Co.), Similac Isomil Advance (soy formula) used to test for differences between MF-fed and BF-fed in- (Ross Products Division, Abbott Laboratories, Columbus, fants and between SF-fed and MF-fed infants to test the 2 hy- Ohio), or Similac Isomil (milk formula) (Ross Products Di- potheses established before conducting the study. It should vision). Formulas were provided as concentrates, and the use also be noted that 2-way ANOVA using feeding group and of each brand name was approximately equal across respec- as the groups was also performed with SigmaStat (version tive groups. The study was approved by the Institutional Re- 3.5, Systat Software, Inc., San Jose, California), and 2-way in- view Board of the University of Arkansas for Medical teractions were considered with comparisons of nominal cat- Sciences, and informed consent was obtained from parents. egory organ characteristics computed with a c2 statistic. A t For the 24 months starting in November 2005, anthropo- test and ANOVA detected the same treatment effects on or- metric measures (weight, length, and head circumference) gans, but in view of the hypotheses being tested, we report were obtained on 4-month-old participants using standard- here P values from t tests only. ized methods, and body composition was assessed by air dis- placement phlethysmography (ADP, PeaPod, Life Measurement, Inc., Concord, California). Participants who Results consented to an ultrasound assessment were studied in the approximate order as they were recruited, and participants No statistically significant group differences were observed entered each diet group at approximately the same rate. for socioeconomic status, race, and maternal anthropomet- This resulted in 18 to 23 boys and girls per diet group who rics (data not shown). Although serum isoflavones were were evaluated in the Radiology Department of the Arkansas not obtained in the participants of this report, mean Children’s Hospital by registered diagnostic medical sonog- (Æ SD) genistein concentrations of 3-month-old infants col- raphers using a sonograms protocol established during a pilot lected during the same time period ranged from 0.59 Æ 0.3 to

216 Gilchrist et al February 2010 ORIGINAL ARTICLES

Table I. Sample characteristics by feeding group and sex BF MF SF Boys Girls Boys Girls Boys Girls n 202018231920 Gestation (wk)* 39.6 Æ 0.2 39.8 Æ 0.3 39.8 Æ 0.3 39.1 Æ 0.2 39.3 Æ 0.2 40.0 Æ 0.3 (kg) 3.7 Æ 0.1 3.6 Æ 0.1 3.6 Æ 0.1 3.4 Æ 0.1 3.5 Æ 0.1 3.4 Æ 0.1 Birth length (cm) 52.3 Æ 0.4 51.3 Æ 0.4 51.9 Æ 0.6 50.5 Æ 0.4 51.3 Æ 0.6 51.1 Æ 0.3 Age at US (mo) 4.2 Æ 0.1 4.0 Æ 0.1 4.3 Æ 0.2 4.3 Æ 0.1 4.2 Æ 0.1 4.1 Æ 0.0 Weight at US (kg)† 6.9 Æ 0.2 6.4 Æ 0.1 7.2 Æ 0.1 6.6 Æ 0.1 6.8 Æ 0.1 6.5 Æ 0.1 Length at US (cm) 63.8 Æ 0.6 61.8 Æ 0.4 63.8 Æ 0.6 62.1 Æ 0.5 63.0 Æ 0.2 62.3 Æ 0.5

US, Ultrasound. *Gestation length is significantly greater in BF than MF girls (P = .03) and in SF than MF girls (P = .01). †Weight at 4 months of age for boys was significantly higher in MF than SF (P = .04).

0.91 Æ 0.3 mg/mL. None of these SF infants had detectable feeding group differences were observed within these normal equol. These genistein values are similar to those reported values. This raises questions about what magnitude of differ- by Setchell et al.16,17 ence would be clinically significant or biologically relevant, A total of 120 infants had a sonographic assessment, includ- especially in light of the relatively high interindividual vari- ing 40 BF (20 boys, 20 girls), 41 MF (18 boys, 23 girls), and 39 ability observed. This cannot be answered by this small clin- SF (19 boys, 20 girls). As shown in Table I, there were no feed- ical study and will require large numbers of study ing group differences in gestational age at birth; birth weight/ participants and a longitudinal study to determine if any length/age; or weight/length at sonography. Body weights of early observations reflect permanent or transient conditions MF boys were greater (P < .05) than of SF boys. and/or if they result in beneficial or adverse health effects. Among girls, there were no significant feeding group dif- This issue of what constitutes an appropriate control for ferences in breast bud or uterine volume, although the formulas is important. In the present report, we presented mean breast volume was slightly lower in SF infants (Figure). Uterine length was <4.0 cm in all participants and did not differ by feeding group (BF: 2.77 Æ 0.31; MF: 2.88 Æ 0.43; SF: 2.88 Æ 0.46.). There were also no significant feeding group differences in the presence of a visible endometrial stripe or shape (heart vs cylindrical vs pear) of the uterus (Table II). MF girls had greater (P < .05) mean ovarian vol- ume than did BF girls, whereas SF infants had a mean ovarian volume that did not differ from BF (Figure). As shown in Table III, the majority of ovaries imaged contained at least 1 follicle per cyst, and there were no feeding group differences in the numbers of mean size of the cysts (Table III). How- ever, MF and SF infants had a greater (P < .01) mean number of follicles per cyst per than did BF infants. Among boys, there were no feeding group differences in breast bud or prostate volumes (Figure). Mean testicular vol- umes did not differ between SF and MF boys, but both were lower (P < .04) than BF boys (BF: 0.855 Æ 0.07; MF: 0.770 Æ 0.03; SF: 0.694 Æ 0.042). No significant diet effects were de- tected in infant weight, length, or body surface area (Table I). No differences were found in body composition (data not shown). Results did not differ when reanalyzed after correct- ing organ volume for any of these variables (data not shown).

Discussion

We prospectively examined the potential relationship be- tween infant diet and hormone-sensitive organ size. Sonogra- Figure. Organ volumes as determined by ultrasound. Bars phy is well tolerated and is a valid methodology to assess 3 13,18-21 represent the mean (ÆSEM) organ volume (cm /kg body wt) organ volumes in small infants. for 4-month-old infants who were BF (20 boys and 20 girls), The observed organ volumes fall within previously pub- MF (18 boys and 23 girls), or SF (19 boys and 20 girls). lished age-appropriate reference norms; however, minor

Ultrasonographic Patterns of Reproductive Organs in Infants Fed Soy Formula: Comparisons to Infants Fed Breast Milk 217 and Milk Formula THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 156, No. 2 comparisons for all 3 diet groups. However, based on the AAP statement of May 2008 that ‘‘human milk is the ideal Table III. Characteristics of ovarian cysts by feeding source of nutrition for infant feeding’’ and the AAP recom- group mendations that breast-feeding be considered the standard Ovaries imaged Ovaries with Cysts per ovary Size of cysts 3 for infant feeding followed by milk-based infant formula, (n)* cysts (n) (n) (cm ) the most appropriate comparisons may be between BF and BF 38 29 1.55 Æ 0.21 0.51 Æ 0.05 MF 42 37 2.84† Æ 0.29 0.46 Æ 0.04 MF and between MF and SF. MF formula varied (P < .05) † SF 36 33 2.64 Æ 0.33 0.43 Æ 0.03 from BF on 3 measures: ovarian volume, numbers of cysts per ovary, and testicular size. Thus, although MF has a long *Ovarian images from 1 SF and 4 MF infants were not used because of artifacts. †P < .01 compared with BF. track record as safe and efficacious, it may have developmen- tal effects that differ from breast milk. On the other hand, the organ volumes of SF infants were the same as BF and/or MF. The organs examined in the current study have been In no case did we observe a difference between infants fed BF demonstrated to be sensitive to exposure. Norma- and SF that was not also reflected in MF infants. This suggests tive ultrasound data of uterine volumes conducted among that SF tracks with either breast milk or milk formula with prepubertal girls from birth to age 15 years suggest that ex- respect to all the effects presented in this study: growth, de- posure to maternal/placental estrogen in utero results in velopment, body composition, and ultrasonically deter- uterine enlargement at birth, which quickly resolves over mined volumes of reproductive organs under the the first month of life and remains low until near the onset conditions of this study. Although we conducted this study of .21,24 However, Trotter et al25 demonstrated that in infants at age 4 months, the period of greatest isoflavone uterine width, depth, and length could be increased postna- exposure for SF infants and an age at which it has been hy- tally during a double-blinded, placebo-controlled study of pothesized that isoflavones would exert biological activity estrogen and replacement among premature in infants,16,17 no evidence of estrogenic effects of SF on infants. Furthermore, these treatment effects per- reproductive organs was found. sisted after the therapy was discontinued. Estrogen exposure Our results differ from 2 previous reports in human in- also results in a ‘‘heart-shaped’’ uterus, in which the fundus fants that used a physical examination to evaluate the effects is larger than the cervix and in the presence of an endome- of SF on breast development. Even though we observed a sex trial echo, which can be observed on sonography. We ob- difference in the volume of breast buds, with girls having served no feeding group differences in uterine length, larger volumes than boys, we do not have an explanation volume, shape, or prevalence of an endometrial stripe. All for this difference. This is in contrast to Bernbaum et al,6 of our measurements were within published reference values who found no sex difference in diameter of breast buds of in- for healthy normal infants. fants by comparing the physical feel of breast buds with a cus- Two groups have previously reported that although ovar- tom set of beads. Their technique would be comparable with ian size increases with pubertal stage, the uterus begins to in- the clinical use of Prader orchidometer beads for assessment crease in size sooner, suggesting that it may be a more of testicular size, which has been shown to be significantly less sensitive indicator of hormonal exposure.24,26 Similar to pre- accurate than ultrasound assessment when compared with vious reports, we observed a high prevalence of ovarian folli- volume measurement of excised organs.20,22,23 Using similar cles per cyst (average 86%) among our sample of healthy methods, female SF infants were reported to have a higher neonates. This did not differ by feeding group, nor did the av- prevalence of breast buds at age 2 years than female infants erage size of follicles per cyst or presence of macrocysts >1 who were fed BF or MF.12 In that study, breast buds were des- cm. However, we did observe larger ovarian volume and ignated to be present if a mass had a diameter >1.5 cm. Inves- more cysts per ovary among female MF infants compared tigators were not blinded to treatment groups in either of with BF infants. Similar to what we have observed in our an- these studies.11,12 Although we were able to assess the pres- imal experiments (unpublished results), there were no statis- ence of breast buds of each child in our study by using ultra- tical differences in ovarian volume between the two formula sound, none of the infants in our study had a breast bud groups (MF vs SF) or between SF and BF groups. The reason diameter >1.5 cm, and no feeding group differences were for the difference between MF and BF is unclear but probably detected in either boys or girls. does not represent an estrogenic effect. Among boys, we observed no feeding group differences in prostate volume, which is -dependent.27 All of our Table II. Uterine characteristics by feeding group volumes were within normative data for boys from 0 to 10 Uterine shape years of age.27 This is similar to what has previously been re- Uterus Visible endometrial Pear Cylindrical Heart ported in animal models of soy exposure, in which no effect imaged (n)* stripe n (%) n (%) n (%) n (%) of SF has been observed on prostate weight.28,29 Although BF 20 17 (85.0) 13 (65.0) 5 (25.0) 2 (10.0) there were no differences in testicular volume between SF MF 19 13 (68.4) 12 (63.2) 4 (21.1) 3 (15.8) and MF boys, both formula groups were significantly lower SF 19 16 (84.2) 9 (47.4) 7 (36.8) 3 (15.8) than BF and were within published normative ranges.11 *Uterine images from 4 MF and 1 SF infants were not used because of artifacts. These results are in direct contradiction to what was observed

218 Gilchrist et al February 2010 ORIGINAL ARTICLES by Tan et al28 who reported an increase in testicular weight 3. Committee on Nutrition. Soy Protein-based formulas: recommenda- and elevated Sertoli and Leydig cell counts among SF-fed tions for use in infant feeding. Pediatrics 1998;101:148-53. male marmoset monkeys compared with their MF-fed male 4. Bhatia J, Greer F. Use of soy protein-based formulas in infant feeding. Pediatrics 2008;121:1062-8. twins. This disparity may be due to the high levels of equol 5. Hilakivi-Clarke L, Cho E, Clarke R. Maternal genistein exposure mimics 30 in infant monkeys. Ultrasound-derived reference data for the effects of estrogen on mammary gland development in female mouse testicular volume among boys from 0 to 6 years suggest offspring. Oncol Rep 1998;5:609-16. that testicular volume increases in the first 5 months of life 6. Ishimi Y, Arai N, Wang X, Wu J, Umegaki K, Miyaura C, Takeda A, and then reaches a prepubertal low at around age 9 months. Ikegami S. Difference in effective dosage of genistein on bone and uterus in ovarectomized mice. Biochem Biophys Res Commun 2000;274: The volume increase coincides with a period of ‘‘mini pu- 697-701. berty’’ resulting from gonadotrophic reaching 7. Awoniyi CA, Roberts D, Veeramachaneni DN, Hurst BS, Tucker KE, a peak at age 3 to 4 months.18 Schlaff WD. Reproductive sequelae in female rats after in utero and neo- There are some limitations to the present study design. natal exposure to the phytoestrogen genistein. Fertil Steril 1998;70: One limitation is that we conducted sonographic assessments 440-7. 8. Casanova M, You L, Gaido KW, Archibeque-Engle S, Janszen DB, at only one age. In light of observed feeding group differ- Heck HA. Developmental effects of dietary phytoestrogens in Sprague- ences, it would be beneficial to have measurements of the Dawley rats and interactions of genistein and daidzein with rat estrogen same children throughout the first 6 to 9 months of life to receptors a and b in vitro. Toxicol Sci 1999;51:236-44. see if there is an actual difference in maximum testicular vol- 9. Santell RC, Chang YC, Nair MG, Helferich WG. Dietary genistein exerts ume or simply a shift in the growth curve due to timing of the estrogenic effects upon the uterus, mammary gland and the hypotha- lamic/pituitary axis in rats. J Nutr 1997;127:263-9. surge. Another limitation is the length of SF ex- 10. Strom BL, Schinnar R, Ziegler EE, Barnhart KT, Sammel MD, posure, which varied between 2 and 4 months. However, this Macones GA, et al. Exposure to soy-based formula in infancy and endo- SF exposure reflects the exposure pattern of infants in the crinological and reproductive outcomes in young adulthood. JAMA United States, as most infants start SF after having problems 2001;286:807-14. with BF or MF. Furthermore, the period between ages 2 and 4 11. Bernbaum JC, Umbach DM, Ragan NB, Ballard JL, Archer JI, Schmidt- Davis H, et al. Pilot studies of estrogen-related physical findings in in- months results in the highest circulating concentrations of fants. Environ Health Perspect 2008;116:416-20. the potentially estrogenic isoflavones. Thereafter, infants 12. Zung A, Glaser T, Kerem Z, Zadik Z. Breast development in the first 2 start complementary foods, and the isoflavone levels begin years of life: an association with soy-based infant formulas. J Pediatr Gas- to decline and the potential effects would be expected to troenterol Nutr 2008;46:191-5. lessen. Thus, if SF effects were to occur, we would expect 13. Estroff JA, Parad R, Stroehla B, Umbach D, Rogan W. Developing methods for studying estrogen-like effects of soy isoflavones in infants that the infant at age 4 months would have the most signifi- 3: Ultrasound. Pediatric Academic Society Annual Meeting Abstract cant effects. 2006;2757:8. We tentatively conclude that anthropometric development 14. Study of Estrogen Activity and Development (SEAD) Soy Study. Avail- and body composition do not differ between formula-fed able at: http://www.niehs.nih.gov/research/atniehs/labs/epi/studies/ and breast-fed infants at age 4 months. Our data do not sup- sead/index.cfm 15. Lambert B. The frequency of mumps and of mumps orchitis and the port major diet-related differences in reproductive organ size consequences for sexuality and . Acta Genet Stat Med 1951;2: as measured by ultrasound in infants at this age. However, 1-166. there is some evidence that ovarian development may be ad- 16. Setchell KD, Zimmer-Nechemias L, Cai J, Heubi JE. Exposure of infants vanced in MF-fed infants compared with BF and SF infants. to phyto-oestrogens from soy-based infant formula. Lancet 1997;350: Furthermore, there was some evidence that formula-fed in- 23-7. 17. Setchell KD, Zimmer-Nechemias L, Cai J, Heubi JE. Isoflavone content fants may have smaller testes and that the numbers of ovarian of infant formulas and the metabolic fate of these phytoestrogens in early cysts per ovary are greater than in BF infants. There is no ev- life. Am J Clin Nutr 1998;68:1453S-61S. idence from the current study that feeding soy formula exerts 18. Kuijper EA, van Kooten J, Verbeke JI, van Rooijen M, Lambalk CB. Ul- any recognizable estrogenic effects on reproductive organs trasonographically measured testicular volumes in 0- to 6-year-old boys. studied. Last, the numbers of infants per diet group were Hum Reprod 2008;23:792-6. 19. Srinivasan M, Aalinkeel R, Song F, Mitrani P, Pandya JD, Strutt B, small and thus the data should be interpreted with caution. n et al. Maternal hyperinsulinemia predisposes rat for hyperin- sulinemia, and -onset obesity and maternal mild food restric- Submitted for publication May 13, 2009; last revision received Jul 9, 2009; tion reverses this phenotype. Am J Physiol Endocrinol Metab accepted Aug 19, 2009. 2006;290:E129-34. Reprint requests: Dr Thomas M. Badger, Arkansas Children’s Nutrition Center, 20. Sakamoto H, Saito K, Oohta M, Inoue K, Ogawa Y, Yoshida H. Testic- 15 Children’s Way, Little Rock, AR 72202. E-mail: [email protected]. ular volume measurement: comparison of ultrasonography, orchidom- etry, and water displacement. Urology 2007;69:152-7. 21. Griffin IJ, Cole TJ, Duncan KA, Hollman AS, Donaldson MD. Pelvic References ultrasound measurements in normal girls. Acta Paediatr 1995;84: 536-43. 1. Merritt RJ, Jenks BH. Safety of soy-based infant formulas containing iso- 22. Taskinen S, Taavitsainen M, Wikstrom S. Measurement of testicu- flavones: the clinical evidence. J Nutr 2004;134:1220S-4. lar volume: comparison of 3 different methods. J Urol 1996;155: 2. National Toxicology Program-Center for the Evaluation of Risks to Hu- 930-3. man Reproduction. NTP-CERHR Expert Panel Report on the Repro- 23. Paltiel HJ, Diamond DA, Di Canzio J, Zurakowski D, Borer JG, Atala A. ductive and Developmental Toxicity of Soy Formula. 1–214. 2006. Soy Testicular volume: comparison of orchidometer and US measurements Formula-06. Ref Type: Report. in dogs. Radiology 2002;222:114-9.

Ultrasonographic Patterns of Reproductive Organs in Infants Fed Soy Formula: Comparisons to Infants Fed Breast Milk 219 and Milk Formula THE JOURNAL OF PEDIATRICS  www.jpeds.com Vol. 156, No. 2

24. Haber HP, Mayer EI. Ultrasound evaluation of uterine and ovarian size 28. Tan KA, Walker M, Morris K, Greig I, Mason JI, Sharpe RM. Infant feed- from birth to puberty. Pediatr Radiol 1994;24:11-3. ing with soy formula milk: effects on puberty progression, reproductive 25. Trotter A, Maier L, Kohn T, Bohm W, Pohlandt F. Growth of the uterus function and testicular cell numbers in marmoset monkeys in adult- and mammary glands and vaginal cytologic features in extremely prema- hood. Hum Reprod 2006;21:896-904. ture infants with postnatal replacement of and progesterone. 29. Perry DL, Spedick JM, McCoy TP, Adams MR, Franke AA, Cline JM. Am J Obstet Gynecol 2002;186:184-8. Dietary soy protein containing isoflavonoids does not adversely affect 26. Salardi S, Orsini LF, Cacciari E, Bovicelli L, Tassoni P, Reggiani A. Pelvic the reproductive tract of male cynomolgus macaques (Macaca fascicula- ultrasonography in premenarcheal girls: relation to puberty and sex hor- ris). J Nutr 2007;137:1390-4. mone concentrations. Arch Dis Child 1985;60:120-5. 30. Gu L, House SE, Prior RL, Fang N, Ronis MJ, Clarkson TB, et al. Meta- 27. Xia SJ, Xu XX, Teng JB, Xu CX, Tang XD. Characteristic pattern of hu- bolic phenotype of isoflavones differ among female rats, pigs, monkeys, man prostatic growth with age. Asian J Androl 2002;4:269-71. and women. J Nutr 2006;136:1215-21.

50 Years Ago in THE JOURNAL OF PEDIATRICS Anticipatory Guidance in Pediatric Practice Hill LF. J Pediatr 1960;56:299-307

n 1960 pediatricians were first and foremost infectious disease specialists. Haemophilus influenzae meningitis and Iepiglottitis, Streptococcus pneumoniae meningitis and pneumonia, Neisseria meningitides disease, newborn sepsis, and other life- and limb-threatening infectious diseases were the battleground that were daily fought by our colleagues. The articles published in pediatric journals at that time were concerned with these infectious diseases, the treatment of premature infants, and trauma. Hill’s prescient discussion about anticipatory guidance is an early look at what has become the main focus of primary care pediatrics in the 21st century. The conquering of the deadly infectious diseases with the Haemophilus influenzae type B, pneumococcal, and meningococcal vaccines, as well as our older routine viral and bacterial , has altered the primary role of the pediatrician from an interventionist to a specialist in prevention. In his discussion, Hill reviews the state of what he, for the first time, labeled ‘‘anticipatory guidance.’’ From normal infantile crying, through development with stranger anxiety, body curiosity, eating patterns, temper tantrums, toilet training, and home safety, his article presages many of the anticipatory guidance issues that primary care pro- viders currently discuss. The need for adolescent independence and discussions about puberty, sexual activity, and sexually transmitted diseases were all discussed 50 years ago in The Journal. The American Academy of Pediatrics along with the Health Resources and Services Administration and other state and federal groups has developed and promoted the ‘‘Bright Futures’’ endeavor. Within its mission lie the core antic- ipatory guidance issues recommended for pediatric practitioners to discuss at routine visits. Hill’s issues will be found prominently within the pages, albeit in more detail. The list of discussion points within the program have progressed far beyond Hill’s dreams and include issues that were unknown at his time. From sleeping position and sudden infant syndrome to sexual identity and HIV prevention, from physical punishment to domestic violence, from foreign adoption issues to gang activity, gun safety and conduct disorder, from seatbelts to car seats, pediatricians must discuss a myriad of topics that are critically important to the healthy outcomes of our children. Although the list is longer, although many of the topics are newer and perhaps less comfortable to discuss, although the focus of the pediatric visit may have changed, Hill accurately and presciently described the importance of leading our parents and children through the labyrinth of development toward a healthy adulthood.

Eugene R. Hershorin, MD Division of General Pediatrics Department of Pediatrics University of Miami Miller School of Medicine Miami, Florida 10.1016/j.jpeds.2009.09.038

220 Gilchrist et al