NATALIE MESTEL NSD648

DIETETICS IN THE LIFESPAN

Infant Soy Formulas

Background: Safety Concerns

Soy formula contains phytoestrogens, including isoflavones. Potential negative effects: Sexual development and reproduction Neurobehavioral development Immune function Thyroid function

Conversely, epidemiologic studies have suggested protective effect of isoflavones against adult chronic diseases, including CHD, breast, endometrial, and prostate cancers [2]

Infants go through developmental stages sensitive to estrogens. Infants more likely to be vulnerable to effects of phytoestrogens in soy [1]

AAP committed to use of human milk as ideal source of nutrition for infants [2]

NIH Addresses Soy Formula 2010NTP (National Toxicology Program) and expert panel used five-level scale to express conclusions: likelihood of adverse human health effect resulting from exposure soy infant formula [2]

o Serious Concern o Concerno Some Concerno Minimal Concerno Negligible

Concern

Use of Soy Protein-Based Formulas in Infant Feeding

(2008) [2]

Updates 1998 AAP review of SF and addresses ongoing concern of phytoestrogens

With term infants, although SF may provide nutrition for normal growth and development, few indications for use in place of CMF. Indications include: (a) infants with galactosemia and hereditary lactase deficiency

For infants with documented cow milk protein allergy, extensively hydrolyzed protein formula should be considered, because 10% to 14% of these infants will also have soy protein allergyo Allergy is reported more often CMF (1.8% - 3.4% vs 0.5% - 1.1%)o Hydrolyzed formula: protein has been broken down so immune system less

likely to recognize it as allergen SF may be indicated when secondary lactose intolerance occurs

(b) in situations in which a vegetarian diet is preferred

SF has no advantage over CMF, unless infant has 1 of indications noted previously

Despite limited indications, SF accounts for about 20% of formula market in U.S [2]

SF composition meets FDA and AAP guidelines for vitamin, mineral, and electrolyte content and provides 67 kcal/dL: Protein: Soy isolate protein, L-methionine, L-carnitine, and taurine, 2.45-2.8 g/100

kcals Fat: mostly vegetable oils, 5.02-5.46 g/100 kcals Carbohydrate: corn maltodextrin, corn syrup solids, and sucrose, 10.26-10.95 g/100

kcals SF contain 20% more calcium and phosphorus than CMF and are fortified with iron

and zinc

No clear evidence that soy isoflavones have adverse effects on human development, reproduction, or endocrine function: Soy isoflavones have different absorption, distribution, metabolism, and excretion

according to age, sex, and cultural group Male infants: no increased feminization Infants with hypothyroidism fed with soy formula have prolonged elevation of

thyroid-stimulating hormone

What does primary literature say???

Use of Soy Protein-Based Formulas in Infant Feeding

(2008) [2]

Body Fat and Bone Mineral Content of Infants Fed Breast Milk, Cow's Milk Formula, or Soy Formula During the First

Year of Life (2013) [3]

Study Design: Prospective Cohort Study

Setting & Sampleo Infants assessed at age 3, 6, 9, and 12 monthso N=207, 84.5% Caucasian, 62% boys, 100%

appropriate for gestational age at birtho No significant differences among 3 feeding

groups in terms of race, sex, gestational age, birth weight, birth length, or SES (Table II)

Measures/Variableso To characterize growth, fat mass (FM), fat-free

mass (FFM), and bone mineral content (BMC) in healthy infants fed breast milk (BM), cow's milk formula (CMF), or soy formula (SF) during 1st year of life

o Standard anthropometric techniques for growth, dual-energy X-rays for body comp.

o Mixed-effects models with repeated measures adjusted for race, SES, gestational age, birth weight, birth length, sex, age, and diet history

Results

① Infants fed BM had higher FM at age 3 months, and lower FFM at age 6-12 months (P < .001)

② Infants fed SF had greater FFM at age 6 months and 9 months compared with infants fed CMF (P < .001)

③ BMC was higher in infants fed BM and lower in infants fed SF at age 3 months (P < .001), but by age 12 months, BMC was significantly higher in infants fed SF

①

①

②

③

③

Results

① Infants fed BM had higher FM at age 3 months, and lower FFM at age 6-12 months (P < .001)

② Infants fed SF had greater FFM at age 6 months and 9 months compared with infants fed CMF (P < .001)

③ BMC was higher in infants fed BM and lower in infants fed SF at age 3 months (P < .001), but by age 12 months, BMC was significantly higher in infants fed SF

Implicationso CMF and SF infants significantly different fat and bone growth trajectories, all

formula fed infants significantly different BM infants o SF infants: leaner body phenotype during 1st year, lower bone mineralization at 3

months, greater bone mineral growth during 1st year compared with infants fed BM or CMF

o Implications for long-term health outcomes and bone health remain unclear

Strengthso Use of precise and accurate technology to measure body comp. and BMCo Future findings may offer insight into long-term consequences of early effects b/c

will be followed up to age 6 years

Weaknesses Unclear how 3 body comp. phenotypes predict later body comp. outcomes in later

life Inconsistent results in other studies due to technological differences 84.5% Caucasian, 62% male Observational cohort, not a randomized control trial

Parents chose which diet to feed infants before study Study could have benefited from having 3 diet groups start respective diets at

birth

Body Fat and Bone Mineral Content of Infants Fed Breast Milk, Cow's Milk Formula, or Soy Formula During the First

Year of Life (2013) [3]

Developmental Status of 1-Year-Old Infants Fed Breast Milk, Cow’s Milk Formula, or Soy Formula (2012) [4]

Study Design: Prospective Cohort Study

Setting & Sampleo N = 391, age 1-2 months, assessed at 3, 6, 9, and 12

months o Mothers: No diagnoses or meds known to affect

fetal/infant development during pregnancies, nonsmokers, no alcohol and no use of soy products during pregnancy or lactation

o Infants all term (≥37 weeks), avg. 6-9 lbs at birth, no medical diagnoses, no meds

o Other exclusion criteria: change of formula after age 2 months and before age 12 months, complementary foods before 4 months, body weight at 3 months <5 kg (11lbs)

Measures/Variableso To characterize mental, motor, and language

developmental status of BF, CMF, or SF infants during 1st year of life

o Development evaluated using Bayley Scales of Infant Development and Preschool Language Scale-3

o Mixed effects models used while adjusting for SES, mother's age and IQ, gestational age, gender, birth weight, head circumference, race, age, and diet history

Bayley Scales:Preschool Language Scale-3:

MDI (Mental Development Index) and PDI (Psychomotor Development Index) assess motor skills, language, and cognitive development at 0-3 years [5] Children presented with tasks

designed to produce observable set of behavioral responses assessed on scales: Cognitive Scale (91 items) Language Composite Scale (97

items) with Receptive and Expressive Language subscales

Motor Composite Scale (138 items) with Fine and Gross Motor subscales

Parent questionnaire with two scales: Social-Emotional and Adaptive Behavior

Assesses receptive and expressive language skills in infants and young children considered to be language precursors using scales [6]: Attention Semantics Structure Integrative Thinking Vocal Development Social Communication

Bayley Scales of Infant Development and Preschool Language Scale-3

①

②

②

③

Results

① MDI (Mental Developmental Index) scores: BF infants slightly higher than SF and CMF at 6-12 months (P < .05)

② PDI (Psychomotor Development Index scores): BF infants higher than SF infants at age 6 months and higher

③ PLS-3 (Preschool Language Scale-3) SF scores higher than CMF infants at 3 and 6 months (P < .05)

Behavioral standardized scores for BF (solid circles), MF (open circles), and SF (inverted triangles) infants during the first year of life.

Results

o Generally, no differences found between CMF versus SF.

o BF consistently scores better for MDI and PDI compared with SF and CMF infants.

Implicationso All scores on developmental testing were within established normal rangeso SF scored within normal limits on standardized testing and did not differ from CMFo Breastfed infants have a slight advantage on cognitive development compared with

CMF and SF

Strengthso Findings similar to other studies: comparable cognitive development between CMF

and SF   Results similar after adjusting for confounding factors

o Large sample size of carefully characterized infants at 4 time points during 1st year of life

Weaknesseso Differences between BF, CMF, and SF were small in magnitude and difficult to

interpret in terms of potential clinical relevanceo Results are limited by observational characteristics of studyo Validity of Bayley Scales of Infant Development assessment needs follow-up of study

group at school age

Developmental Status of 1-Year-Old Infants Fed Breast Milk, Cow’s Milk Formula, or Soy Formula (2012) [4]

Early Life Soy Exposure and Age at Menarche (2012) [7]

Study Design: Prospective Cohort Study

Setting & Sample: o N=2920 girlso Subjects participants in Avon Longitudinal

Study of Parents and Children (ALSPAC) in United Kingdom

o Mothers enrolled during pregnancyo Eligible subjects limited to term, white

females

Measures/Variables:o To examine timing of menarche in relation to

effects of soy isoflavone exposure in SFo Infant feeding questionnaires administered

during infancyo Grouped into BM, ECMF, ESF, and LSFo Age at menarche assessed through

questionnaires administered annually between ages 8-14.5

o Kaplan-Meier survival curves and Cox hazards models used to assess age and risk of menarche

Kaplan-Meier Curves Cox Hazard Models

Performs survival analysis Used in study interested in time

until participants present specific event or endpoint In this study: point of menarche

Usually, end of study reached before all participants presented event, and outcome of remaining patients is unknown. Outcome unknown of LTF participants (censored data) [8]

Hazard is a measure of how rapidly event of interest occurs

Can be interpreted as chance of event occurring in treatment arm divided by chance of event occurring in control arm

Endpoints depicted using Kaplan-Meier survival curves Curve represents odds of endpoint

having occurred at each point in time (hazard)

Hazard ratio is relationship between hazards in two groups and represents distance between Kaplan-Meier plots [8]

Kaplan-Meier Survival Curves and Cox Proportional Hazards Models

Number of subjects living at start – Number of subjects died

Number of subjects living at startS=

Results

o 2% mothers reported soy products introduced to at or before 4 months (Early Soy, ESF)

o Median age at menarche: 153 months, 12.8 years

o Median age at menarche for ES girls: 149 months, 12.4 years

o Compared to ECMF, ESF girls at 25% higher risk of menarche throughout follow up (Hazard Ratio 1.25 [95% CI, 0.92, 1.71])

Implicationso Girls fed ESF may have increased risk of early menarcheo Association needs more in-depth evaluation for future studyo Replication of findings is imperative before meaningful conclusions can be drawn

Strengthso Analysis used large, longitudinal cohort study that is generalizable to U.K.o Maternal age, education, breastfeeding duration, and child health not associated

with age at menarche, so should not biaso Maternal age at menarche and prenatal smoking adjusted to control for bias related

to LTFo Survival analysis showed subtle link between SF and menarche not found in

previous studies

Weaknesseso Few subjects (54 girls) in ESF, only Caucasiano Unclear if formula products in U.K. differed from in U.S.: results can not be

generalizedo Exact soy intake in sample unknown, so true dose relationship could not be assessedo ECMF and LSF differed demographics: potential for confounding in these groups

relative to ECMF and ESFo Selection bias in final study sample: excluded girls had lower birth weight, shorter

breast feeding duration, more likely to be ill as infants, exposed to prenatal tobacco smoke, and born to younger, heavier mothers.

Early Life Soy Exposure and Age at Menarche (2012) [7]

References

[1] http://www.niehs.nih.gov/health/topics/agents/sya-soy-formula/ [2] Bhatia MD, Jatinder, Greer MD, Frank, Committee on Nutrition. (2008). Use of Soy

Protein-Based Formulas in Infant Feeding. Pediatrics. 121:5, 1062 -1068  [3] Andres, Aline PhD., Casey, Patrick H. MD., Cleves, Mario A. PhD., Bader, Thomas M.

PhD., (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. The Journal of Pediatrics, 163,1: 49-54

[4] Andres, Aline PhD., Cleves, Mario A. PhD., Bellando, Jayne B. PhD., Pivik, R.T. PhD., Casey, Patrick H. MD., Bader, Thomas M. PhD., (2012). Developmental Status of 1-Year-Old Infants Fed Breast Milk, Cow’s Milk Formula, or Soy Formula. Pediatrics: Official Journal of the American Academy of Pediatrics, 129,6: 1134-1140.

[5] http://archive.acf.hhs.gov/programs/opre/ehs/perf_measures/reports/resources_measuring/res_meas_cdif.html

[6] http://clas.uiuc.edu/special/evaltools/cl02539.html [7] Adgent, Margaret., Daniels, Julie., Rogan, Walter., Adair, Linda., Edwards, Lloyd.,

Westreich, Daniel., Maisonet, Mildred., Marcus, Michele. (2012) Early Life Soy Exposure and Age at Menarche. National Institutes of Health. 26,2: 163-175

[8] http://en.wikipedia.org/wiki/Kaplan%E2%80%93Meier_estimator

Conclusions and Questions to Consider

AAP states term infants fed with soy formula have normal growth and development, energy intake, serum albumin, and bone mineralization; there are no clinical problems with nutrition, sexual development, thyroid disease, or immune function [2]

Even with no stated advantage over CMF, if primary articles, NIH, and AAP generally express safety of soy formulas…

Why does AAP state there are few indications for SF use in place of CMF?

Why is SF only recommended for allergy to CMF and dietary preference and not for everyday use along with CMF?

Is there discrepancy between the primary literature and public recommendations?