SUPPLEMENT ARTICLE

Infant Feeding Practices and Reported Food Allergies

at 6 Years of Age
AUTHORS: Stefano Luccioli, MD,a Yuanting Zhang, PhD,b
Linda Verrill, PhD,b Moraima Ramos-Valle, MS,a and Ernest abstract
Kwegyir-Afful, PhDa
OBJECTIVE: The goal of this study was to identify the frequency of
Offices of aFood Additive Safety, and bAnalytics and Outreach,
physician-diagnosed food allergies among 6-year-old US children
Center for Food Safety and Applied Nutrition, Food and Drug
Administration, College Park, Maryland and study the impact of exclusive breastfeeding and complementary
KEY WORDS food introduction on this frequency.
breastfeeding, children, complementary foods, food allergy, METHODS: Data were analyzed from children who participated in the
physician diagnosis, prevalence, risk factors
Infant Feeding Practices Study II Year 6 Follow-Up Study (Y6FU).
ABBREVIATIONS
Children with probable food allergy (pFA) were defined as children
aOR—adjusted odds ratio
IFPS II—Infant Feeding Practices Study II with report of physician-diagnosed food allergy at age 6 years.
pFA—probable food allergy Subgroups of pFA included children who were not diagnosed before
Y6FU—Year 6 Follow-Up Study 1 year of age (new pFA) and those with atopic risk factors (high risk).
Dr Luccioli conceptualized and designed the study, directed the
analyses, and drafted the initial manuscript; Drs Zhang and
RESULTS: Prevalence of total pFA in the Y6FU was 6.34%. The majority of
Kwegyir-Afful gathered the data, designed and conducted the these children had new pFA and high-risk factors. Higher maternal
initial analyses, and reviewed and revised the manuscript; and education, higher family income, family history of food allergy, and
Dr Verrill and Ms Ramos-Valle co-designed the analyses and
reported eczema before 1 year of age were significantly associated
critically reviewed the manuscript. All authors approved the
final manuscript as submitted. with higher odds of total or new pFA. Exclusive breastfeeding duration
The findings and conclusions in this report are those of the and timing of complementary food introduction were not significantly
authors and do not necessarily represent the views of the US associated with total pFA. However, exclusive breastfeeding of $4
Food and Drug Administration. months compared with no breastfeeding was marginally associated
www.pediatrics.org/cgi/doi/10.1542/peds.2014-0646E with lower odds of new pFA (adjusted odds ratio: 0.51; P = .07); this
doi:10.1542/peds.2014-0646E effect was not observed with high-risk children.
Accepted for publication May 20, 2014 CONCLUSIONS: Analysis of infant and maternal variables in the Y6FU
Address correspondence to Stefano Luccioli, MD, Office of Food cohort of US children revealed that socioeconomic and atopic factors
Additive Safety, Center for Food Safety and Applied Nutrition, Food
were the main predictors of pFA at age 6 years. Exclusive breastfeeding
and Drug Administration, 5100 Paint Branch Parkway, HFS 200,
College Park, MD 20740. E-mail: stefano.luccioli@fda.hhs.gov of $4 months may have a preventive effect on development of pFA after
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). 1 year of age in non high-risk children. Pediatrics 2014;134:S21–S28
Copyright © 2014 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have
no financial relationships relevant to this article to disclose.
FUNDING: This study was funded by the US Food and Drug
Administration, Centers for Disease Control and Prevention,
Office on Women’s Health, National Institutes of Health, and
Maternal and Child Health Bureau in the US Department of
Health and Human Services.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated
they have no potential conflicts of interest to disclose.

PEDIATRICS Volume 134, Supplement 1, September 2014 S21

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
Food allergies are a recognized public old from the IPFS II have previously been a physician because of a food allergy
health problem. Epidemiologic evidence reported.18 The present article expands was available for 1531 children. If the
has shown that the prevalence of food on the previous article by using the mother responded affirmatively, she was
allergies in US children has increased longitudinal data to characterize the asked whether the child was diagnosed
over the past decade1,2 and that de- development of food allergies in the IFPS by a physician as having an allergy to any
velopment of food sensitization in early II sample when the children are 6 years food. Mothers were also asked whether
life may predispose individuals to later old. Our main objective was to assess the 6-year-old had outgrown a food al-
allergic conditions (eg, asthma) in child- the predictors of pFA in this age group, lergy or intolerance that he or she had
hood.3,4 The increased incidence of food particularly the impact of feeding prac- had when younger. The final analytical
allergy cannot be explained by ge- tices such as breastfeeding and timing sample was further restricted to 1363
netic factors alone, and studies linking of complementary food introduction at children by removing those with missing
food allergies to other emerging health important early life dietary milestones values for any variables used in the re-
conditions in children involving diet (eg, development (ie, at 4 and 6 months, re- gression model. Analysis of the 179 ex-
obesity, diabetes)5,6 suggest that dietary spectively) on food allergies. cluded children showed no significant
factors may be important. Thus, it is differences from the final analytical
necessary to better understand how METHODS sample with regard to prevalence of pFA
and whether certain dietary choices, at ages 1 and 6 years and variables re-
Study Design
practices, or related environmental lated to high risk for atopy.
The IFPS II was a mail panel survey
causes early in life are contributing to
conducted in 2005–2007 of US mothers
the development of food allergies in Dependent Variables
and their newborn infants. Survey data
young children. The main dependent variable was total
were collected approximately monthly
There have been recent controversies from pregnancy through the infant’s pFA (all children with a current physi-
over recommended feeding practices first 12 months of life and queried var- cian diagnosis of food allergy at age 6
during infancy for the prevention of ious information on infant feeding and years). Additional dependent variables
food allergies and other allergic con- health, including food allergies, mater- included new pFA (subset of children
ditions in later life.7–9 Although most nal diet and health, and demographic with physician diagnosis of food allergy
pediatric organizations continue to and environmental variables.17 Eligible at age 6 years but with no diagnosis
promote exclusive breastfeeding for at mothers enrolled in IFPS II were re- before 1 year of age) to study the po-
least the infant’s first 6 months of life,10,11 contacted in 2012 to provide follow-up tential for effects during the infancy
the optimal timing for introduction of information for themselves and their 6- period on later food allergic manifes-
complementary foods has not been year old child. The Y6FU survey queried tations and high-risk pFA (subset of
well established. There is inconclusive various information about the child children with pFA at age 6 years and
evidence that either early introduction such as current health status, including report of any of the following atopic risk
before 4 months of age increases, or physician diagnosis of food allergy, food factors: family history of food allergy,
delayed introduction after 6 months of environment, home environment, and family history of other atopy, or eczema
age reduces, the risk for subsequent other related outcomes. The study was before age 1 year).9,20
food allergic sensitization, especially approved by the US Food and Drug
for children with atopic risk factors.9,12–15 Administration’s Research Involving Hu- Independent Variables
The Infant Feeding Practices Study II man Subjects Committee. Details on the Infant feeding variables were derived
(IFPS II) and the Year 6 Follow-Up Study methods and sample characteristics of from IFPS II data. Exclusive breastfeeding
(Y6FU), a longitudinal survey of .1500 participants in Y6FU are described in duration was calculated as the midpoint
pregnant women that assessed com- another article in this supplement.16 between the infant’s age when the
prehensive dietary and health histories mother last reported feeding only breast
of their children from birth through 12 Sample milk and the age when the mother first
months and again at 6 years of age, The total sample size of the Y6FU study reported not exclusively breastfeeding.21
offer opportunities to study early life was 1542 children, representing 52% of Of note, neonates reportedly given for-
practices that contribute to food al- the original IFPS II population who were mula in the hospital but exclusively
lergy development by age 6 years.16,17 eligible for Y6FU.19 Response by moth- breastfed after discharge were consid-
Certain characteristics of a probable ers to the Y6FU survey question asking ered exclusively breastfed. The timing of
food allergy (pFA) in infants ,1 year if their child had ever been taken to introduction of solid or liquid foods was

S22 LUCCIOLI et al
Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
 SUPPLEMENT ARTICLE

estimated by using the midpoint be- formula or complementary food intro- a possible food allergy, 37.2% reported
tween the age at which the food was not duction was assessed at time periods that they had outgrown a food allergy or
reported as being fed to the infant over based on important dietary milestones of intolerance that they had had when
the previous 7 days and the age at which either 4 or 6 months. Because only 9 younger.
food was first reported. Two main food mothers reported exclusive breastfeeding Total (n = 1363) and high-risk (n = 823)
categories were defined: milk allergen– for $6 months, 4 months was used as the population groups were identified by
derived foods, including formula (the cutoff point for exclusive breastfeeding using logistic regression to analyze and
majority of which [.95%] was milk duration. Because exclusive breastfeeding compare frequencies of pFA and other
based), cow’s milk, other milk (eg, goat), and milk/formula introduction were col- population characteristics or differences
and dairy (eg, cheese); and comple- linear events, logistic regression results in relation to various demographic, en-
mentary foods, comprising mostly solid were presented for only the exclusive vironmental, and health variables (Ta-
foods such as any soy, fish/shellfish, breastfeeding group. Analysis of individual ble 1). The subpopulation of children
peanut, eggs, infant cereals, other cereals, complementary foods produced incon- reported to have persistent pFA was too
fruit, vegetables, French fries, meats, clusive results because the numbers were small to provide meaningful results and
and sweet foods (eg, pudding). too small to provide a stable model for the therefore was not analyzed separately. It
Background demographic and health in- various time points. Thus, we instead was found that 89% (79 of 89) of children
formation came from the Y6FU and IFPS II counted the earliest time for the in- with pFA had at least 1 atopic factor.
databases, with missing Y6FU demo- troduction of any of the complementary There were no significant differences in
graphic data imputed from data in IFPS II. foods in year 1. If the foods were not characteristics between the high-risk
Y6FU demographic information included reported, the individual data were counted and total populations or with regard to
child’s gender and mother’s education, as “not reported.” Because complemen- pFA frequency in relation to exclusive
race, and household income (#185% vs tary foods were analyzed as a group, breastfeeding duration, timing of com-
.185% of the federal poverty level). Y6FU a record was removed from the analysis plementary food introduction, or other
health variables included whether the only if data for the entire group were variables.
mother or anyone in the family had a his- missing. To understand the relative impact
of individual allergenic food groups, cross- Predictors of pFA
tory of food allergy or other atopy (ie,
asthma, eczema, environmental aller- tabulations of timing of introduction of Relevant demographic, environmental,
gies). IFPS II data included whether the these foods in relation to frequency of pFA and health factors were assessed for
child was delivered vaginally or by ce- (ie, percent with pFA) at each respective association with total pFA and among the
sarean delivery and whether the child milestone was constructed. The x 2 analy- new pFA or high-risk subgroups by using
had physician-diagnosed food allergy or ses were conducted to compare the dif- logistic regression analysis (Table 2).
reported eczema at any time before age 1 ferences within each food group. For both total and new pFA samples,
year. The smoking exposure variables controlling for all predictors, the ad-
were constructed from both Y6FU data on RESULTS justed odds of having pFA at age 6 years
current smoking status of the mother and were nearly double (adjusted odds ratio
Prevalence and Characteristics of [aOR]: 1.86; P = .01) among children who
other household members and IFPS II
pFA had a family history of food allergy than
data on whether mother and others had
smoked in the house during pregnancy or From the total population of Y6FU among those who did not. Children with
the child’s first year of life. The number of respondents, 97 (6.34%) of 1531 mothers pFA were more likely to have mothers
siblings was constructed from both IFPS II reported a physician diagnosis of food who had higher education (especially
(number of other siblings in household) allergy in their child at age 6 years; by college/bachelor’s degree) or were
and Y6FU (number of other siblings born comparison, 3.89% of total children in from families with higher income (aOR:
since IFPS II) data. the IFPS II year 1 survey had a reported 1.70; P = .06) compared with those with
food allergy diagnosis. Among children incomes ,185% of the poverty level.
Statistical Analysis with pFA by age 6 years, 78 children The highest odds associated with total
SAS version 9.3 (SAS Institute, Inc, Cary, developed a newly diagnosed pFA since 1 or new pFA were found for children with
NC) was used forall analyses. Because all yearofage(ie, hadpFA at year6 butnotat reported eczema before age 1 year
dependent variables are binary, logistic year 1). Only 19 had evidence of per- (aOR: 3.69; P , .001). For the high-risk
regression was used. The association sistent pFA (ie, had pFA in both years 1 subsample, the main predictor was in-
between pFA and duration of exclusive and 6). Among children who had ever come .185% of the poverty level (aOR:
breastfeeding or timing of any milk/ been taken to a physician because of 3.19; P = .002).

PEDIATRICS Volume 134, Supplement 1, September 2014 S23

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
TABLE 1 Descriptive Statistics and Percent Frequency of pFA for Total, New pFA, and High-Risk Exclusive Breastfeeding Duration
Groups
and Timing of Complementary Food
Variable Total Groupa (n = 1363) High-Risk Groupc
Introduction
(n = 823)

n (%) % pFA % New pFAb n (%) % pFA

As shown in Table 1, significant predictors
of pFA were similar between total and
pFA 89 7.0 5.2 (n = 71) 79 9.6
Mother’s education new pFA except for exclusive breastfeed-
High school 180 (13.2) 2.7 1.7 101 (12.3) 5.0 ing variables. Exclusive breastfeeding
Associate degree 174 (12.8) 6.1 4.6 95 (11.5) 9.5 duration for $4 months compared with
Some college 348 (25.5) 5.4 6.8 204 (24.8) 7.8
Bachelor 470 (34.5) 8.7 6.8 300 (36.4) 12.7
no exclusive breastfeeding was associ-
Postgraduate 191 (14.0) 6.8 4.6 123 (15.0) 8.9 ated with marginally significant lower
Race odds for new pFA (aOR: 0.51; P = .067) but
Non-Hispanic white 1177 (86.4) 6.5 5.3 708 (86.0) 9.5
not for total pFA or pfA in the high-risk
Non-Hispanic Asian 26 (1.9) 7.7 7.7 15 (1.8) 13.3
Non-Hispanic black 53 (3.9) 5.7 3.8 35 (4.2) 8.6 population (Table 2). These findings sug-
Hispanic 77 (5.6) 7.7 6.5 42 (5.1) 14.3 gest that exclusive breastfeeding for
Non-Hispanic other 30 (2.2) 3.3 23 (2.8) 4.3 a minimum of 4 months may offer pro-
Income/percent federal
poverty level tection from developing food allergies in
,185% 529 (38.8) 4.3 2.4 319 (38.8) 6.6 the postinfancy period but only for chil-
$185% 834 (61.2) 7.9 7.0 504 (61.2) 11.5 dren not at high risk for food allergy.
Child’s gender
Female 675 (49.5) 5.6 4.6 394 (47.9) 8.1 The timing of any complementary food
Male 688 (50.5) 7.4 5.8 429 (52.1) 11.0 introduction was not independently
Parity
associated with total, new, or high-risk
No sibling 142 (10.4) 9.2 8.4 85 (10.3) 14.1
1 sibling 555 (40.7) 7.0 5.9 325 (39.5) 9.5 pFA at defined dietary milestones (Ta-
$2 siblings 666 (48.9) 5.6 3.9 413 (50.2) 8.7 ble 2). Table 3 displays the feeding
Type of delivery patterns of selected allergenic foods at
Vaginal 959 (70.4) 5.8 4.7 592 (71.9) 8.4
Cesarean 404 (29.6) 8.2 6.4 231 (28.1) 12.6 respective milestones and comparative
Family history of food allergy pFA frequencies for the total and high-
No 1053 (77.3) 5.6 4.6 513 (62.3) 9.6 risk samples. Milk-based foods and
Yes 310 (22.7) 10.0 7.4 310 (37.7) 9.7
Family history of other atopy
cereals, and to a lesser extent fruits and
No 764 (56.0) 5.2 5.0 224 (27.2) 13.4 vegetables, were the foods most com-
Yes 599 (44.0) 8.2 5.5 599 (72.8) 8.2 monly introduced before 6 months of
Reported eczema before
age, with ,2% of infants introduced to
age 1 y
No 1082 (79.4) 4.3 3.6 542 (65.9) 6.8 allergenic foods such as egg, peanuts,
Yes 281 (20.6) 15.0 11.4 281 (34.1) 15.0 fish, or soy before 6 months. In both
Maternal tobacco smoke samples, consistently lower pFA were
exposure
No 1162 (85.2) 6.4 5.2 700 (85.0) 9.1 observed in children fed milk, egg, or
Yes 201 (14.8) 7.5 5.0 123 (15.0) 12.2 peanuts by 12 months of age compared
Other tobacco smoke exposure with those not reported to be fed these
in home
No 1189 (87.2) 6.5 5.2 725 (88.1) 9.2
foods; the inverse association was
Yes 174 (12.8) 6.9 5.2 98 (11.9) 12.2 noted for soy, fruits, and infant cereals.
Exclusive breastfeeding duration However, no statistically significant dif-
0 mo 590 (43.3) 7.4 6.1 345 (41.9) 11.3
ferences were found.
1–3 mo 411 (30.2) 5.8 5.1 249 (30.3) 8.4
$ 4 mo 362 (26.5) 5.8 3.9 229 (27.8) 8.3
Complementary food introduction DISCUSSION
by age of infant
1–3 mo 461 (33.8) 7.4 5.6 288 (35.0) 11.1 From the nationally distributed IFPS II
4–5 mo 542 (39.8) 6.3 5.4 322 (39.1) 8.7 sample of .1500 children participat-
6–12 mo 236 (17.3) 6.8 4.6 141 (17.1) 10.6
Not reportedd 124 (9.1) 4.0 4.0 72 (8.8) 5.6 ing in the Y6FU study, we analyzed
a Includes all Y6FU children included in the logistic regression analysis. cases of maternally reported physician
b Includes children within the total group who had pFA at age 6 years but did not have pFA at age 1 year. diagnoses of food allergy to study the
c Includes subset of total group children with any of the following: family history of food allergy, family history of other atopy,

or reported eczema before 1 year of age.

frequency of pFA in 6-year-old US chil-
d Includes children not reported to have complementary foods introduced during the IFPS II year 1 survey. dren and to describe present-day and

S24 LUCCIOLI et al
Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
 SUPPLEMENT ARTICLE

TABLE 2 aOR Estimates for pFA in Total, New pFA, and High-Risk Groups as a Function of early life predictors of food allergies in
Background Characteristics and Feeding Practices
this age group. We found a prevalence
Variables Total pFAa New pFAb High-Risk pFAc of pFA of 6.34% in this sample, which is
aOR 95% CL aOR 95% CL aOR 95% CL within the 3.9% to 8% range of national
Mother’s education prevalence rates for children estimated
(High school)d 1 — 1 — 1 — by using parental reports of food aller-
Associate degree 2.1 0.71–7.01 2.15 0.58–10.26 1.74 0.42–8.72
Some college 1.89 0.73–5.86 2.6 0.83–11.43 2.16 0.66–9.77
gies in North American children.20,22–24
Bachelor 2.95* 1.16–9.12 3.15† 1.03–13.83 3.18† 1.02–14.08 The majority (89%) of children with pFA
Postgraduate 1.89 0.63–6.49 2.38 0.67–11.25 2.13 0.58–10.29 had at least 1 atopic risk factor. In this
Race
(Non-Hispanic white) 1 — 1 — 1 —
sample, we also observed that pFA
Non-Hispanic Asian 0.99 0.15–3.72 1.08 0.17–4.09 1.48 0.22–6.20 reported at age 6 years was diagnosed
Non-Hispanic black 0.74 0.16–2.25 0.55 0.09–2.03 0.83 0.13–3.08 after age 1 year in at least 75% of chil-
Hispanic 1.49 0.54–3.48 1.48 0.49–3.68 1.97 0.63–5.19
Non-Hispanic other 0.33 0.02–1.75 NE — NE —
dren and, in one-third or more of chil-
Income/percent of federal dren who had ever been taken to
poverty level a physician for a possible food allergy,
(,185%) 1 — 1 — 1 —
a previous food allergy or intolerance
$185% 1.70† 0.99–3.00 2.53** 1.32–5.16 3.19** 1.57–6.99
Child’s gender had resolved by age 6 years. Using lo-
(Female) 1 — 1 — 1 — gistic regression to control for a variety
Male 1.17 0.74–1.86 1.15 0.70–1.92 1.31 0.76–2.29
of confounding variables, higher mater-
Parity
(No sibling) 1 — 1 — 1 — nal education and family income, family
1 sibling 0.81 0.41–1.68 0.69 0.34–1.48 0.58 0.26–1.32 history of food allergy, and reported
$2 siblings 0.78 0.39–1.66 0.56 0.26–1.26 0.56 0.25–1.32 eczema before age 1 year were the most
Type of delivery
(Vaginal) 1 — 1 — 1 — significant predictors of pFA at age 6
Cesarean 1.37 0.84–2.21 1.31 0.76–2.21 1.23 0.67–2.21 years. These findings support other
Family history of food allergy observations showing the relative im-
(No) 1 — 1 — — —
Yes 1.86* 1.12–3.03 1.90* 1.08–3.28 — — portance of socioeconomic and atopic
Family history of other atopy factors in childhood food allergies.22,25–27
(No) 1 — 1 — — — Early life feeding practices, such as
Yes 1.26 0.80–2.01 0.89 0.53–1.50 — —
Reported eczema before exclusive breastfeeding duration and
age 1 y timing of introduction of complementary
(No) 1 — 1 — — — foods in relation to important dietary
Yes 3.69** 2.31–5.91 3.47** 2.05–5.86 — —
Maternal tobacco smoke milestones at 4 or 6 months of age, were
exposure not significantly associated with overall
(No) 1 — 1 — 1 — pFA. However, children who were exclu-
Yes 1.26 0.60–2.51 0.96 0.40–2.10 1.1 0.45–2.51
Other tobacco smoke exposure
sively breastfed for at least 4 months
in home had borderline significantly lower odds
(No) 1 — 1 — 1 — of developing a new pFA compared with
Yes 1.42 0.64–2.98 1.5 0.61–3.39 1.81 0.70–4.34
Exclusive breastfeeding duration
those who were not exclusively breast-
(0 mo) 1 — 1 — 1 — fed. This potential benefit was not ob-
1–3 mo 0.72 0.42–1.23 0.78 0.43–1.38 0.81 0.42–1.51 served in high-risk atopic children.
$ 4 mo 0.69 0.36–1.29 0.51† 0.24–1.03 0.58 0.26–1.25
Complementary food introduction The literature has shown inconsistent
by infant age results regarding the relationship be-
(1–3 mo) 1 — 1 — 1 —
tween prolongedexclusivebreastfeeding
4–5 mo 0.83 0.47–1.45 0.98 0.53–1.80 0.91 0.46–1.77
6–12 mo 0.93 0.45–1.86 0.87 0.37–1.89 0.96 0.40–2.20 and food allergy in children.9,15,28 For
Not reportede 0.64 0.21–1.60 0.84 0.28–2.29 0.69 0.19–1.95 studies that have shown a preventive
CL, confidence limits; NE, not established due to insufficient numbers. —, denote logistic regression reference variables for benefit, the benefit seems short-lived
background characteristics, feeding practices and high risk group. †P , .10, *P , .05, **P , .01.
a Includes all children who had pFA at age 6 years. (up to a few years of age) and limited
b Includes children who had pFA at age 6 years but did not have pFA at age 1 year.
to atopic children with specific food al-
c Includes children with pFA at age 6 years and any of the following: family history of food allergy, family history of other atopy,

or reported eczema before age 1 year.

lergies (ie, cow’s milk).15,28–30 Our results
d Reference categories are indicated by parentheses. instead seem to show a breastfeeding
e Includes children not reported to have complementary foods introduced during the IFPS II year 1 survey.

PEDIATRICS Volume 134, Supplement 1, September 2014 S25

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
TABLE 3 Number of Children Introduced to Milk Allergen–Derived and Allergenic Complementary Foods in Total and High-Risk Samples at Respective
Time Points and Percent Frequency of pFA
Sample Food Introduced Milk Allergen–Derived Complementary Foods: Select Individual Allergenic Groupsa
Foods
Egg Peanut Soy Fish Fruit Vegetable Cerealsb
Total group,c % pFA = 6.5 N 1363 1259 1315 1343 1325 1176 1177 1200
,4 mo Yes 899 5 4 3 1 119 82 451
(% pFA) (6.7) (0.0) (0.0) (0.0) (0.0) (8.4) (3.6) (7.3)
4–5 mo Yes 59 6 2 7 1 535 585 484
(% pFA) (6.8) (0.0) (0.0) (14.0) (0.0) (6.2) (7.4) (5.8)
6–12 mo Yes 300 617 254 96 209 438 423 176
(% pFA) (5.3) (5.5) (4.7) (9.4) (5.7) (6.6) (6.4) (6.8)
Not reported Yes 105 631 1055 1237 1114 84 87 89
(% pFA) (8.6) (7.6) (7.1) (6.4) (6.6) (6.0) (5.8) (4.5)
High-risk group,d % pFA = 9.6 n 823 760 801 815 801 714 715 729
,4 mo Yes 530 2 3 2 0 75 55 280
(% pFA) (10.0) (0.0) (0.0) (0.0) (0.0) (13.3) (5.4) (11.0)
4–5 mo Yes 42 5 2 3 1 325 350 291
(% pFA) (10.0) (0.0) (0.0) (33.3) (0.0) (9.2) (10.8) (8.2)
6–12 mo Yes 180 377 147 61 124 266 261 109
(% pFA) (7.8) (8.2) (6.1) (14.8) (9.0) (9.8) (9.6) (11.0)
Not reported Yes 71 376 649 749 676 48 49 49
(% pFA) (11.3) (11.4) (10.8) (9.2) (9.0) (6.2) (6.1) (6.1)
a Data do not include frequencies of complementary foods: meats, French fries, and sweet foods.
b Data include infant cereals and other cereals.
c Includes all children at age 6 years included in the logistic regression analysis.
d Includes children with family history of food allergy or other atopy or reported eczema before age 1 year.

benefit for nonatopic children and are missed. Moreover, our study focused on is not nationally representative and
consistent with the Tasmanian Asthma children with pFA at age 6 years, not on therefore is limited in ascribing preva-
Study,31 which, adjusting for familial risk other current allergic conditions (ie, lence estimates to the US population.19
factors of maternal, paternal, or sibling eczema) for which protective or negative Third, our prevalence data on the pFA
atopy, found reduced odds of reported effects of complementary food intro- group do not discern whether the
physician-diagnosed food allergies in 7- duction in infancy have been shown.33,34 reported probable food allergies were
year-old children who were exclusively Future analyses of the relationships immunoglobulin E mediated or could
breastfed for $3 months. between breastfeeding, formula feeding, have represented other food allergic
With a focus on defined 4- and 6-month and complementary food type and in- disorders in early childhood, such as
dietary milestones, results from the pres- troduction in relation to children with celiac disease and eosinophilic gastro-
ent study concur with findings from a combination of allergic conditions may intestinal diseases.20 Moreover, analysis
a German birth cohort, which found no be warranted to fully evaluate the as- of food allergies in this study was not
significantoveralleffectofcomplementary/ sociation between infant feeding practi- linked to any specific food. Thus, it is not
solid food feeding practices on food ces and prevention of food allergies or possible to assess the potential impact
allergic sensitization at age 6 years.14 It other allergic diseases. of early or delayed introduction of re-
should be noted, however, that children The present study has some important spective complementary foods on pro-
in the IFPS II study cohort were infants limitations. First, the criteria for select- moting or preventing individual food
during the years 2005 to 2007, a time ing the pFA group were not based on allergies. Fourth, the impact on pFA
period when prevailing advice to US clinically validated methods but from frequency by other potentially relevant
mothers was to delay introduction of parentally reported cases of physician- and confounding factors tied to the in-
food allergens.32 Indeed, a very small diagnosed food allergy. The accuracy cidence of atopic disorders, such as
percentage (,2%) of children were of this method in estimating the true environment (eg, rural farm versus ur-
reported to have been fed non-milk prevalence of clinical food allergy is ban living, infectious exposures)35 and
allergens (nuts, eggs, fish/shellfish, or unknown and may be subject to parent allergic sensitization between ages 1
soy) before 6 months of age. The relative or physician biases, especially in cases and 6 years, is missing from this logistic
impact of this delayed introduction of in whichthe diagnosiswasmade without regression analysis.
allergenic foods on our study’s mile- diagnostic testing. Second, although Despite these limitations, the main
stone end points cannot be easily dis- nationally distributed, the IFPS II sample strengths of the Y6FU study are the large

S26 LUCCIOLI et al
Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
 SUPPLEMENT ARTICLE

sample size of .1500 children, the in- group, are particularly relevant to ascer- exclusive breastfeeding of $4 months
clusion of questions that address many tain how factors in early life relate to was marginally associated with lower
different pFA-associated factors likely to development of pFA later in life. odds of developing pFA at age 6 years.
affect outcomes, and the ability to exam- This potential benefit was not observed
ine the association between information CONCLUSIONS among the high-risk atopic children,
on infant factors, including feeding before In this cohort of 6-year-old US children, which suggests the need to separate
age 1 year and reported physician- socioeconomic (higher maternal edu- children according to atopic risk when
diagnosed pFA by age 6 years in more cation and income) and atopic (family studying preventive benefits of exclusive
than one-half of the original participants history of food allergy and infant eczema) breastfeeding on food allergy.
from the IFPS II study. The longitudinal factors were significant predictors of
study design with a short time frame over pFA. Ouranalysis did not find a significant ACKNOWLEDGMENT
which mothers were asked to recall in- association between pFA and feeding The authors thank the supplement edi-
formation in early infancy, as well as the practices at established dietary mile- tors, Mary Ditto, Steven Gendel, Karl
additional detailed questions relative to stones in infancy. However, among chil- Klontz, and Jordan Lin, for their critical
most population surveys in this age dren who did not have pFA by age 1 year, review of this manuscript.

REFERENCES
1. Branum AM, Lukacs SL. Food allergy among 9. Greer FR, Sicherer SH, Burks AW; American 15. Muraro A, Dreborg S, Halken S, et al. Di-
U.S. children: trends in prevalence and hos- Academy of Pediatrics Committee on Nutri- etary prevention of allergic diseases in
pitalizations. NCHS Data Brief. 2008;10:1–8 tion; American Academy of Pediatrics Section infants and small children. Part III: critical
2. Sicherer SH, Muñoz-Furlong A, Godbold JH, on Allergy and Immunology. Effects of early review of published peer-reviewed observa-
Sampson HA. US prevalence of self- nutritional interventions on the development tional and interventional studies and final
reported peanut, tree nut, and sesame al- of atopic disease in infants and children: the recommendations. Pediatr Allergy Immunol.
lergy: 11-year follow-up. J Allergy Clin role of maternal dietary restriction, breast- 2004;15(4):291–307
Immunol. 2010;125(6):1322–1326 feeding, timing of introduction of comple- 16. Fein SB, Li R, Chen J, Scanlon KS, Grummer-
3. Schroeder A, Kumar R, Pongracic JA, et al. mentary foods, and hydrolyzed formulas. Strawn LM. Methods for the year 6 follow-
Food allergy is associated with an in- Pediatrics. 2008;121(1):183–191 up study of children in the Infant Feeding
creased risk of asthma. Clin Exp Allergy. 10. Section on Breastfeeding. Breastfeeding Practices Study II. Pediatrics. 2014;134(suppl 1):
2009;39(2):261–270 and the use of human milk. Pediatrics. S4–S12
4. Brockow I, Zutavern A, Hoffmann U, et al; 2012;129(3). Available at: www.pediatrics. 17. Fein SB, Grummer-Strawn LM, Raju TN. In-
GINIplus Study Group. Early allergic sensi- org/cgi/content/full/129/3/e827 fant feeding and care practices in the
tizations and their relevance to atopic 11. Agostoni C, Decsi T, Fewtrell M, et al; United States: results from the Infant
diseases in children aged 6 years: results ESPGHAN Committee on Nutrition. Comple- Feeding Practices Study II. Pediatrics. 2008;
of the GINI study. J Investig Allergol Clin mentary feeding: a commentary by the 122(suppl 2):S25–S27
Immunol. 2009;19(3):180–187 ESPGHAN Committee on Nutrition. J Pediatr 18. Luccioli S, Ross M, Labiner-Wolfe J, Fein SB.
5. Caffarelli C, Cavagni G, Pierdomenico R, Gastroenterol Nutr. 2008;46(1):99–110 Maternally reported food allergies and
Chiari G, Spattini A, Vanelli M. Coexistence 12. Nwaru BI, Takkinen HM, Niemelä O, et al. other food-related health problems in
of IgE-mediated allergy and type 1 diabetes Introduction of complementary foods in infants: characteristics and associated
in childhood. Int Arch Allergy Immunol. infancy and atopic sensitization at the age factors. Pediatrics. 2008;122(suppl 2):S105–
2004;134(4):288–294 of 5 years: timing and food diversity in S112
6. Visness CM, London SJ, Daniels JL, et al. a Finnish birth cohort. Allergy. 2013;68(4): 19. Fein SB, Labiner-Wolfe J, Shealy KR, Li R,
Association of obesity with IgE levels and 507–516 Chen J, Grummer-Strawn LM. Infant Feed-
allergy symptoms in children and adoles- 13. Poole JA, Barriga K, Leung DY, et al. Timing ing Practices Study II: study methods. Pe-
cents: results from the National Health and of initial exposure to cereal grains and the diatrics. 2008;122(suppl 2):S28–S35
Nutrition Examination Survey 2005-2006. J risk of wheat allergy. Pediatrics. 2006;117 20. Boyce JA, Assa’ad A, Burks AW, et al; NIAID-
Allergy Clin Immunol. 2009;123(5):1163– (6):2175–2182 Sponsored Expert Panel. Guidelines for the
1169, e1–e4 14. Zutavern A, Brockow I, Schaaf B, et al; LISA diagnosis and management of food allergy
7. Cattaneo A, Williams C, Pallás-Alonso CR, Study Group. Timing of solid food in- in the United States: report of the NIAID-
et al. ESPGHAN’s 2008 recommendation for troduction in relation to eczema, asthma, sponsored expert panel. J Allergy Clin
early introduction of complementary foods: allergic rhinitis, and food and inhalant Immunol. 2010;126(suppl 6):S1–S58
how good is the evidence? Matern Child sensitization at the age of 6 years: results 21. Zhang Y, Carlton E, Fein SB. The association
Nutr. 2011;7(4):335–343 from the prospective birth cohort study of prenatal media marketing exposure re-
8. Fewtrell M, Wilson DC, Booth I, Lucas A. Six LISA. Pediatrics. 2008;121(1). Available at: call with breastfeeding intentions, initia-
months of exclusive breast feeding: how www.pediatrics.org/cgi/content/full/121/1/ tion, and duration. J Hum Lact. 2013;29(4):
good is the evidence? BMJ. 2011;342:c5955 e44 500–509

PEDIATRICS Volume 134, Supplement 1, September 2014 S27

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
22. Gupta RS, Springston EE, Warrier MR, et al. lergy: results of the LISA birth cohort study. 31. Matheson MC, Erbas B, Balasuriya A, et al.
The prevalence, severity, and distribution Clin Exp Allergy. 2010;40(3):450–457 Breast-feeding and atopic disease: a cohort
of childhood food allergy in the United 27. Tan RA, Corren J. The relationship of rhi- study from childhood to middle age. J Al-
States. Pediatrics. 2011;128(1). Available nitis and asthma, sinusitis, food allergy, lergy Clin Immunol. 2007;120(5):1051–1057
at: www.pediatrics.org/cgi/content/full/ and eczema. Immunol Allergy Clin North 32. American Academy of Pediatrics. Commit-
128/1/e9 Am. 2011;31(3):481–491 tee on Nutrition. Hypoallergenic infant for-
23. Soller L, Ben-Shoshan M, Harrington DW, 28. van Odijk J, Kull I, Borres MP, et al. mulas. Pediatrics. 2000;106(2 pt 1):346–349
et al. Overall prevalence of self-reported Breastfeeding and allergic disease: a mul- 33. Kumar R, Caruso DM, Arguelles L, et al. Early
food allergy in Canada. J Allergy Clin tidisciplinary review of the literature (1966– life eczema, food introduction, and risk of
Immunol. 2012;130(4):986–988 2001) on the mode of early feeding in food allergy in children. Pediatr Allergy
24. Branum AM, Lukacs SL. Food allergy among infancy and its impact on later atopic mani- Immunol Pulmonol. 2010;23(3):175–182
children in the United States. Pediatrics. festations. Allergy. 2003;58(9):833–843 34. Zutavern A, Brockow I, Schaaf B, et al; LISA
2009;124(6):1549–1555 29. Lucas A, Brooke OG, Morley R, Cole TJ, Study Group. Timing of solid food in-
25. Ben-Shoshan M, Harrington DW, Soller L, Bamford MF. Early diet of preterm infants troduction in relation to atopic dermatitis
et al. Demographic predictors of peanut, and development of allergic or atopic dis- and atopic sensitization: results from
tree nut, fish, shellfish, and sesame allergy ease: randomised prospective study. BMJ. a prospective birth cohort study. Pediat-
in Canada. J Allergy (Cairo). 2012;2012: 1990;300(6728):837–840 rics. 2006;117(2):401–411
858306 30. Saarinen UM, Kajosaari M. Breastfeeding 35. von Mutius E. Allergies, infections and the
26. Schnabel E, Sausenthaler S, Schaaf B, et al; as prophylaxis against atopic disease: hygiene hypothesis—the epidemiological
LISA Study Group. Prospective association prospective follow-up study until 17 years evidence. Immunobiology. 2007;212(6):433–
between food sensitization and food al- old. Lancet. 1995;346(8982):1065–1069 439

S28 LUCCIOLI et al
Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018
 Infant Feeding Practices and Reported Food Allergies at 6 Years of Age
Stefano Luccioli, Yuanting Zhang, Linda Verrill, Moraima Ramos-Valle and Ernest
Kwegyir-Afful
Pediatrics 2014;134;S21
DOI: 10.1542/peds.2014-0646E

Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/134/Supplement_1/S21
References This article cites 33 articles, 12 of which you can access for free at:
http://pediatrics.aappublications.org/content/134/Supplement_1/S21.
full#ref-list-1
Permissions & Licensing Information about reproducing this article in parts (figures, tables) or
in its entirety can be found online at:
https://shop.aap.org/licensing-permissions/
Reprints Information about ordering reprints can be found online:
http://classic.pediatrics.aappublications.org/content/reprints

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it
has been published continuously since . Pediatrics is owned, published, and trademarked by the
American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,
60007. Copyright © 2014 by the American Academy of Pediatrics. All rights reserved. Print
ISSN: .

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018

 Infant Feeding Practices and Reported Food Allergies at 6 Years of Age
Stefano Luccioli, Yuanting Zhang, Linda Verrill, Moraima Ramos-Valle and Ernest
Kwegyir-Afful
Pediatrics 2014;134;S21
DOI: 10.1542/peds.2014-0646E

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/134/Supplement_1/S21

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it
has been published continuously since . Pediatrics is owned, published, and trademarked by the
American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,
60007. Copyright © 2014 by the American Academy of Pediatrics. All rights reserved. Print
ISSN: .

Downloaded from http://pediatrics.aappublications.org/ by guest on January 8, 2018