DavisEarly_Infant_Temperament2013TICN

Top Clin Nutr
Vol. 28, No. 4, pp. 312–323
Copyright c 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins
PEDIATRIC NUTRITION ISSUES
Association of Early Infant
Temperament to Weight Gain
Velocity and Intake in
3-Month-Old Infants
Anne M. Davis, PhD, RD; Jay Segal, PhD
Early excessive infant weight gain is considered an indicator for risk of childhood obesity. The
aim of this study was to determine whether there is an association between infant temperament
and infant formula intake and weight gain. An observational study for temperament was designed
inside a primary 4-month, randomized, double-blind, controlled, and parallel term infant formula
study. The validated Carey Early Infant Temperament Questionnaire was distributed to infants’
mothers at 3 months for assessment. Infants were enrolled in 2 pediatric clinics in the Manila area,
the Philippines. Two hundred twenty-two healthy term newborn infants completed the study.
Multivariate regression models (basic, complete, and maximum) that controlled for potential
confounders found inverse relationship between frequency of formula intake and adaptability
(r2
= − 0.61; 0.285 standard error [SE]; P = .032) and frequency of formula intake and mood
(r2
= − 0.70; 0.259 SE; P = .007). Similarly, inverse relationships were measured between total
daily formula intake and adaptability (r2
= − 160.99; 53.269 SE; P = .003), mood (r2
= − 165.48;
48.442 SE; P = .001), distractibility (r2
= − 137.38; 49.197 SE; P = .006), and manageability
(r2
= − 71.34; 28.45 SE; P = .013). This is the first study to demonstrate an association between
temperament and frequency and amount of infant formula feeding. Long-term effects of these
feeding practices on later eating behaviors and weight gain need to be examined. Key words:
childhood obesity, early infancy, infant temperament, weight gain velocity
Author Affiliations: Department of Nutrition and
Public Health, School of Health and Natural
Sciences, University of Saint Joseph, West Hartford,
Connecticut (Dr Davis); Department of Public
Health, Temple University, Philadelphia,
Pennsylvania (Dr Segal).
The authors gratefully thank William B. Carey, MD
(Children’s Hospital of Philadelphia; author of the
EITQ), for his interest and consulting service to the en-
tire study and translation of the EITQ. The authors are
indebted to Nicole Stouffer, MS (Stouffer & Associates, 8
Wellesley Way, Medford, NJ 08055, and Pfizer [formerly
Wyeth] Nutrition, 200 Campus Drive, Collegeville, PA
19426), for her expertise and diligent statistical contri-
bution and education.
Conflicts of Interest and Source of Funding: Dr Davis
conducted this substudy research as a component of
the infant formula study while employed at Wyeth Nu-
trition as a clinical scientist. The substudy served as
Dr Davis’ doctoral dissertation. Wyeth Nutrition funded
INTERNATIONAL surveys on the global
prevalence of obesity have reached a pan-
demic scale.1
In 2007, it was estimated that
22 million children younger than 5 years or
the entire research study (approximately $2.0 million).
The lead author (former Wyeth employee) was required
to wait until Wyeth published the infant formula study
before publishing the infant temperament substudy.
The authors have disclosed that they have no signif-
icant relationships with, or financial interest in, any
commercial companies pertaining to this article.
Correspondence: Jay Segal, PhD, Department of Pub-
lic Health, Temple University, 1301 Cecil B. Moore
Ave, Ritter Annex 9th floor (004-09), Philadelphia,
PA 19122 (Jay.Segal@temple.edu).
DOI: 10.1097/01.TIN.0000437409.66565.cf
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
312

Association of Early Infant Temperament to Weight Gain Velocity and Intake 313
less were considered overweight2
; now, ap-
proximately 10% to 20% of infants and tod-
dlers in the United States are overweight3
(body mass index [BMI] >85th percentile for
age) and close to 10% of those infants and
toddlers, from birth to 2 years, are also obese3
(ratio of weight to recumbent length ≥95th
percentile for age). Risk factors for the de-
velopment of obesity are multifactorial, inter-
related, and include but are not limited to
genetic, biological, political, societal, cultural,
psychological, and lifestyle factors.3
Identifi-
cation of risk and the strength of infant obe-
sity risk factors are not well established. The
steady increase in the prevalence of childhood
overweight and obesity highlights the urgent
need for the identification of early-life risk
factors and anticipatory guidance in public
health.4,5
Multiple population and cohort studies
have demonstrated associations between in-
fant growth patterns and childhood and adult
obesity.6-13
Stettler and colleagues10
identi-
fied a correlation between rapid weight gain
during the first week of life and higher inci-
dence of overweight in adulthood. Similarly,
Stettler et al in another study found a correla-
tion between rapid infant weight gain within
the first 3 months of life and later develop-
ment of adult obesity.9
In a systematic review
of 21 studies, Ong and colleagues14
reported
that rapid weight gain during infancy was the
constant factor associated with increased sub-
sequent obesity risk. Ekelund et al9
showed
that the risk for metabolic syndrome was pre-
dicted by rapid weight gain during infancy
(0–6 months).
Although rapid early infant weight gain has
been shown to predict later risk for obesity,
it is unclear what factor(s) is (are) driving the
surplus intake that leads to excessive infant
weight gain.15-18
A possible theory is the po-
tential risk of infant temperament trait(s), and
a parent’s response to infant temperament
consequently influences additional infant in-
take and weight gain. The role of tempera-
ment traits and weight gain has limited re-
search in healthy young infants. Lucas and
St. James-Roberts19
reported on young infants
who cried most frequently also gained the
most weight. Temperament has been asso-
ciated with the frequency of crying/fussing
in infants at 6 weeks of age.20
Clinical stud-
ies have implicated rapid weight gain in fussy
older infants and slower weight gain in “easy”
older infants.21
OBJECTIVE
The aim of this observational infant temper-
ament study was to determine whether there
is a relationship between infant temperament
style and infant formula intake and/or infant
weight gain. The assumption is that when an
infant temperament is difficult, it will predict
that the infant demands more attention in-
cluding additional feeding for nonnutritional
purposes (eg, soothing and calming), which
in turn leads to excess weight gain. The pur-
pose of the primary study was to evaluate in-
fant growth and safety of a new term infant
formula with added lutein.22
METHODS
Setting
Nested within the primary prospective, ran-
domized, controlled, double-blind trial with 2
parallel groups of healthy term infants was
this secondary observational nonrandomized
study planned as an infant temperament ques-
tionnaire. The trial was conducted at 2 study
centers: Asian Hospital Medical Center in
Muntinlupa City and Pula Health Center in
Cabuyao Laguna, the Philippines.
Design and participants
A total of 240 healthy full-term Asian infants
younger than 14 days or less were screened
and enrolled. A total of 116 infants were ran-
domized to each formula group (232 total
infants) (Figure 1), with 110 infants in each
group completing the study. Potential coer-
cion was averted with procedures that were
strictly followed so that mothers were not ap-
proached for study enrollment until (1) after
they were discharged from the hospital and

314 TOPICS IN CLINICAL NUTRITION/OCTOBER–DECEMBER 2013
Figure 1. Study infant disposition.
(2) after the mother reported to have made
the choice to bottle-feed her infant.
The validated Early Infant Temperament
Questionnaire (EITQ) was completed by
mothers of 3-month-old healthy term infants
in the Manila area of the Philippines, with ap-
proximately equal numbers of male and fe-
male infants in the study sample.23
Because
infant behavior and emotion was being mea-
sured, breastfed infants were not included in
the study. Exclusive formula-fed infants were
chosen for this study because it would be dif-
ﬁcult to separate out temperament trait differ-
ences between breastfed and formula-fed in-
fants due to the maternal–infant dyad bonding
of breastfeeding.24,25
The unique and strong
maternal–infant bonding observed in breast-
fed infants, less robust in the formula-fed dyad,
could confound temperament observations
and measurements.26
The primary study formulas [Gold (control
formula) and Gold + Lutein (experimental
formula)] were supplied as ready-to-feed liq-
uid, in 250-mL tetra briks. The 2 study for-
mulas had the same composition of micronu-
trients and macronutrients with the excep-
tion of lutein, which was added at 200 mcg/L
to Gold + Lutein (Table 1). The primary
study formulas had a mean caloric density of
0.67 calories (kcal) per milliliter (Table 1).22
Other micronutrients were identical between
the study formula and the control for-
mula, with the exception of a carotenoid.
Carotenoid values were higher in infants fed
lutein formula.22
The study formula was fed
ad libitum. Study procedures (Table 2) con-
sisted of infant weights taken at baseline, 4, 8,
and 12 weeks and measurements of changes
in weight, weight gain velocity, 3-day weighed
formula intake records during weeks 7 and 11
Table 1. Study Formula Macronutrients
Macronutrient Unit
Gold
Formula
Gold ±
Lutein
Formula
Protein (whey,
casein)
g 14 14
Carbohydrate
(lactose)
g 73 73
Fat g 36 36
Energy kcal 672 672
Arachidonic acid mg 123 123
Docosahexaenoic
acid
mg 71 71
Linoleic acid mg 5800 5800
Linoleic–linolenic
ratio
10:1 10:1
Nucleotides mg 30 30

Table 2. Schedule of Study Proceduresa
aThe text shown in italics denotes clinic visit.
for a total of 6 days of assessed total volume
of formula intake (mL/d), total daily calories
per day, frequency of intake (as the number
of bottle-feedings per day). The 76-item EITQ
was completed by week 12.
The purpose of the temperament ques-
tionnaire data collection tool was to evalu-
ate and categorize infant behavior and emo-
tion to be compared with infant weight gain
and intake.27
The EITQ was translated to
Tagalog for the Philippines for restandardiza-
tion. Analysis was provided for internal consis-
tency and reliability because cultural percep-
tions of infant temperament may differ and
there is no current information regarding the
Philippines.28
Parents completed the infant
temperament questionnaire at 11 weeks and
returned the completed questionnaire to the
investigator at week 12. The 9 traits of infant
temperament (activity, biological rhythmic-
ity, initial approach/withdrawal, adaptability,
intensity, mood, persistence/attention span,
distractibility, and sensory threshold) were
calculated according to the EITQ.29
All infants
were assessed by the mother’s impression of
manageability and the 9 temperament traits.29
Statistical analysis
The sample size was determined from the
primary study design to have sufficiently large
power to exhibit that growth of the Gold
group is equivalent to the Gold + Lutein
group as measured by average weight gain
per day (g/d). Z scores were calculated using
the SAS program offered by the Centers for
Disease Control and Prevention found on:
http://www.cdc.gov/nccdphp/dnpa/growth
charts/resources/sas.htm.30
Besides infant
temperament trait data, the covariate data
collected were an additional 14 independent
variables including maternal education level,
maternal occupation, family income, mater-
nal family support, infant sex, infant’s birth
order, maternal age, prenatal class, prenatal
anxiety or depression, postnatal anxiety or
depression, maternal smoking status, mater-
nal prepregnancy BMI more than 40 kg/m2
,
maternal history of diabetes, parity as number
of live births, infant’s birth order, and the
infant’s weight and z score at baseline. The
relationship between the secondary study
endpoint temperament (the temperament
trait scores and the mother’s impression of
manageability) and the outcome variables
measured at visit week 12 (weight gain [g/d],
daily frequency of intake, and the total daily
intake) was examined using Pearson correla-
tion coefficients and linear regression models.
The intention of the analysis was to examine
the magnitude and significance level of the

linear relationship, identified by the slope,
for each of the temperament variables (the 9
temperament traits and mother’s impression
of manageability) in separate models for each
outcome (weight gain, frequency of intake,
and total daily intake).
A linear relationship measured by the slope
was examined in 3 linear models (basic, com-
plete, and maximum). The basic model only
looked at the temperament traits that had sta-
tistically significant coefficients with the de-
pendent variables (weight gain, frequency of
intake, or total daily intake). The complete
model used the forward selection method to
look at select covariables (maternal education
level, maternal occupation, family income, in-
fant sex, maternal age, prenatal class, prenatal
anxiety or depression, infant’s age at the visit,
the infant’s weight at baseline, and the z score
at baseline) to explain or predict 1 of the 3 de-
pendent variables. The maximum model con-
tained all additional covariates (temperament
traits and covariables) whether they were
significant or not and compared all at one
time.
Ethics approval
The study protocol was reviewed and ap-
proved by the Philippines Ministry of Health,
the Institutional Review Board for the Pulo
Clinic, and the Asian Hospital Institutional
Review Board. In addition, the protocol ex-
ecution, study management, monitoring, con-
fidentiality of data collection and manage-
ment, and protection of human subjects were
done by the primary investigator who was
responsible for complying with the proto-
col and adherence to Good Clinical Practice/
International Conference on Harmonization
guidelines. A Wyeth Nutrition study moni-
tor visited the investigator before the start of
the study for training and at regular intervals
thereafter. All information was recorded on
source documents, and data were recorded in
the case report form screens. Computerized
and manual data edit checks were performed
on all entered data to ensure the data were
valid and uniform.
RESULTS
A total of 232 infants were enrolled and
randomized by computer assignment in the
Manila region of the Philippines. The infant
3-month study dropout rate of 5% was low
(Figure 1). Common adverse events were
formula intolerance and lactose intolerance,
whereas the most common reason for study
withdrawal was parental request, because one
parent not being in agreement with study par-
ticipation.
Infant demographic and baseline character-
istics were collected (Table 3). Some note-
worthy baseline maternal characteristics were
that 43% of the mothers had their first-born
infant in the study (Table 4). None of the
mothers had any history of diabetes or a pre-
natal BMI greater than 40 kg/m2
. Reported
postnatal anxiety or depression was less than
prenatal anxiety or depression (2.3% vs 9.1%,
respectively). The mean maternal education
level was 11.2 ± 1.78 years, and 2.3% or
5 out of 220 mothers smoked during their
pregnancy.
Mean weight gain velocity was 38.33 ±
9.179 g increase per day at 4 weeks, 34.32 ±
7.301 g increase per day at 8 weeks, and
31.29 ± 5.95 g increase per day at 12 weeks.
Mean length gain velocity was 0.146 ± 0.045
cm per day at 4 weeks, 0.128 ± 0.026 cm
per day at 8 weeks, and 0.117 ± 0.02 cm per
day at 12 weeks. Mean head circumference
(HC) gain velocity was 0.069 ± 0.028 cm per
day at 4 weeks, 0.058 ± 0.016 cm per day
at 8 weeks, and 0.053 ± 0.016 cm per day
at 12 weeks. Most infant growth for the total
infants, female infants, and male infants fell
within normal limits on the basis of z scores
for weight-for-age, length-for-age, HC-for-age,
and weight-for-height-for-age z scores and per-
centiles at study visits. Normal limits defined
by the World Health Organization are − 1.88
to 1.88 z scores.31
Mean total daily infant formula intake in-
creased over time from 963.47 ± 330.86 mL
at study week 4 to 1194.9 ± 495.22 mL at
study week 8 to 1253.7 ± 505.12 mL at study
week 12. Infant mean frequency of feeding

Table 3. Summary of Infant Demography
and Baseline Characteristics Population:
Intent to Treat
Total
(N = 220)
Age, d
Mean 9.75
STD 3.49
Median 10
Minimum 1
Maximum 14
Sex
Female 112 (50.9)
Male 108 (49.1)
Race
Asian 220 (100)
Length, cm
Mean 50.1
STD 1.8
Median 50
Minimum 45.5
Maximum 55.5
Weight, kg
Mean 3.19
STD 0.34
Median 3.17
Minimum 2.52
Maximum 4.19
HC, cm
Mean 34.4
STD 1.12
Median 34.5
Minimum 31.3
Maximum 37.6
Weight-for-age z score
Mean − 0.73
STD 0.667
Median − 0.69
Minimum − 2.3
Maximum 1.0
Length-for-age z score
Mean − 0.67
STD 0.880
Median − 0.64
Minimum − 3.1
Maximum 2.0
Abbreviations: HC, head circumference; STD, standard
deviation.
Table 4. Summary of Mother Demography
Intent to Treat
Total
(N = 220)
Mother’s age, yr
Mean 27.6
STD 5.79
Median 27
Minimum 18
Maximum 45
Parity
1 95 (43.2)
2 54 (24.5)
3 37 (16.8)
≥4 34 (15.5)
Birth order of the child
1 95 (43.2)
2 54 (24.5)
3 36 (16.4)
≥4 35 (15.9)
Family income per US capita, $
1000–1999 9 (4.1)
2000–3999 121 (55)
4000–7999 78 (35.5)
8000–16000 12 (5.5)
Family emotional support
No 1 (0.5)
Yes 219 (99.5)
History of diabetes
No 220 (100)
Pregnancy BMI >40 kg/m2
No 220 (100)
Smoked during pregnancy
No 215 (97.7)
Yes 5 (2.3)
Participated in prenatal or parenting class
No 41 (18.6)
Yes 179 (81.4)
Prenatal anxiety or depression
No 200 (90.9)
Yes 20 (9.1)
Postnatal anxiety or depression
No 215 (97.7)
Yes 5 (2.3)
Mother’s/guardian’s education level
N 220
Mean 11.2
(continues)

Table 4. Summary of Mother Demography
Intent to Treat (Continued)
Total
(N = 220)
STD 1.76
Median 10
Minimum 6
Maximum 16
Mother’s/guardian’s occupation
Clerk 16 (7.3)
Homemaker 10 (4.5)
Legislator/official/manager 1 (0.5)
Plant/machine
operator/assembler
53 (24.1)
Professional 34 (15.5)
Service/shop/market sales 53 (24.1)
Skilled agricultural/fishery 8 (3.6)
Technician/associate
professional
21 (9.5)
Abbreviations: BMI, body mass index; STD, standard
deviation.
was 11.54 ± 3.029 bottle-feedings per day at
study week 4, 11.46 ± 3.087 bottle-feedings
per day at study week 8, and 10.75 ± 2.071
bottle-feedings per day at study week 12. The
mean total daily caloric intakes were 645.52
± 221.68 kcal at study week 4, 800.58 ±
331.78 kcal at study week 8, and 839.98 ±
338.43 kcal at study week 12.
Correlations were analyzed by Pearson
correlation coefficient. There were no signifi-
cant relationships found in either of the basic,
complete, or maximum regression models
between weight gain and adaptability or
weight gain and mood (Table 5). Correlations
found were between frequency of daily
intake and adaptability (r = − 0.143; P =
.034), total daily intake and adaptability
(r = − 0.196; P = .003), total daily intake and
distractibility (r = − 0.172; P = .011), total
daily intake and manageability (r = − 0.179;
P = .008), frequency of daily intake and mood
(r = − 0.161; P = .017), and total daily intake
and mood (r = − 0.214; P = .014) (Table 5).
Linear regression is used to model the re-
lationship between temperament trait and
weight gain or total intake or frequency of in-
take and confounders or covariates by fitting
a linear equation to observed data. The tem-
perament trait variable is considered to be the
explanatory variable, and weight gain or total
intake or frequency of intake is considered to
be the dependent variable. The Pearson corre-
lation coefficient determines whether there is
a significant association between 2 variables.
Correlation analysis is often used with regres-
sion analysis because correlation analysis is
used to measure the strength of association.
Regression analysis is used to create a mathe-
matical model that can predict the values of
weight gain or total daily formula intake or
frequency of infant formula intake on the ba-
sis of the temperament trait. In the regression
analysis, data are plotted in 2 dimensions as a
scatter plot. The scatter plot is a visual inspec-
tion of the data prior to see whether the rela-
tionship between the 2 variables is increasing
or decreasing and gives only a rough idea of
the relationship.
There was a significant inverse rela-
tionship found between frequency of in-
take (or the number of bottle-feedings
per day) and adaptability for the basic
(r2
= − 0.59; 0.279 standard error (SE);
P = .034), complete (r2
= − 0.58; 0.274 SE;
P = .036), and maximum (r2
= − 0.61; 0.285
SE; P = .032) regression models. A signifi-
cant inverse relationship was shown between
frequency of intake and mood for the basic
(r2
= − 0.63; 0.262 SE; P = .017), complete
(r2
= − 0.68; 0.255 SE; P = .009), and max-
imum (r2
= − 0.70; 0.259 SE; P = .007) re-
gression models.
There was a significant inverse relationship
found between total daily intake and adapt-
ability for the basic (r2
= − 152.16; 51.614
SE; P = .004), complete (r2
= − 150.47;
51.97 SE; P = .004), and maximum regression
(r2
= − 160.99; 53.269 SE; P = .003) mod-
els. A significant inverse correlation between
total daily intake and distractibility occurred
in the basic (r2
= − 122.21; 47.404 SE; P =
.011), complete (r2
= − 124.98; 47.37 SE;

Table 5. Correlation of Weight Gain, Intake, and Temperament Traits Population: Intent to
Treat (N = 220)
Correlation Coefficient ra
P
Weight Gain to week 12 vs activity − 0.1184 .0796
Frequency of daily intake (bottle-feedings) vs activity − 0.0904 .1813
Total daily intake (mL) vs activity − 0.0710 .2947
Weight gain to week 12 vs adaptability 0.0413 .5424
Frequency of daily intake (bottle-feedings) vs adaptability − 0.1428 .0343b
Total daily intake (mL) vs adaptability − 0.1958 .0035b
Weight gain to week 12 vs approach − 0.0495 .4654
Frequency of daily intake (bottle-feedings) vs approach − 0.0251 .7109
Total daily intake (mL) vs approach − 0.0910 .1789
Weight gain to week 12 vs distractibility − 0.0465 .4929
Frequency of daily intake (bottle-feedings) vs distractibility − 0.0673 .3202
Total daily intake (mL) vs distractibility − 0.1720 .0106b
Weight gain to week 12 vs intensity − 0.1098 .1044
Frequency of daily intake (bottle-feedings) vs intensity 0.0829 .2204
Total daily intake (mL) vs intensity 0.0402 .5533
Weight gain to week 12 vs manageability − 0.0438 .5185
Frequency of daily intake (bottle-feedings) vs manageability − 0.1240 .0664
Total daily intake (mL) vs manageability − 0.1793 .0077b
Weight gain to week 12 vs mood − 0.0466 .4913
Frequency of daily intake (bottle-feedings) vs mood − 0.1611 .0168b
Total daily intake (mL) vs mood − 0.2135 .0014b
Weight gain to week 12 vs persistence 0.0224 .7408
Frequency of daily intake (bottle-feedings) vs persistence − 0.0451 .5062
Total daily intake (mL) vs persistence − 0.0810 .2314
Weight gain to week 12 vs rhythmicity − 0.0041 .9517
Frequency of daily intake (bottle-feedings) vs rhythmicity − 0.0014 .9837
Total daily intake (mL) vs rhythmicity 0.0642 .3435
Weight gain to week 12 vs threshold − 0.0084 .9011
Frequency of daily intake (bottle-feedings) vs threshold 0.0451 .5058
Total daily intake (mL) vs threshold 0.1279 .0583
aPearson correlation coefficient.
bStatistically significant P < .05.
P = .009), and maximum (r2
= − 137.38;
49.197 SE; P = .006) regression models. There
was also an inverse relationship seen between
total daily intake and manageability for the ba-
sic (r2
= − 75.93; 28.24 SE; P = .008), com-
plete (r2
= − 68.38; 27.795 SE; P = .015), and
maximum (r2
= − 71.34; 28.45 SE; P = .013)
regression models. A significant association
found was an inverse relationship between
total daily intake and mood for the basic (r2
=
− 156.21; 48.402 SE; P = .001), complete
(r2
= − 161.29; 47.412 SE; P = .001), and
maximum (r2
= − 165.48; 48.442 SE; P =
.001) regression models.
DISCUSSION
The results of this study adds to the growing
body of evidence suggesting that difficult tem-
perament may contribute to excessive infant
intake as a risk factor of childhood overweight

or obesity. Increasing temperament trait dif-
ficulties in mood, adaptability, manageabil-
ity, and distractibility seems to be related
to increases in total daily intake (mL) and
frequency of intake (the number of bottle-
feedings per day). Similarly, Carey et al32
found that negative mood predicted increased
weight gain in infants 6 to 12 months of
age. In another study, a group of distressed
infants who were difficult to soothe were
found to have higher skinfold thickness than
their nondistressed counterparts.33
Potential
long-term effects of temperament on feed-
ing have been observed in a group of chil-
dren when their mothers rated their chil-
dren to have difficult temperaments, mea-
sured by hyperactivity, unpredictability, and
low attention span. These difficult traits were
then found to be associated with excess
weight gain in middle childhood by Carey
and colleagues.32
Ravaja and Keltikakangas-
Jarvinen34
found that a high level of negative
emotionality and a lack of energy correlated
with higher BMI in boys aged 9 to 15 years.
These aforementioned studies potentially ex-
plain 3% to 16% of weight gain in children.
In this study as well as in others,32-34
more
infants with difficult temperament traits con-
sumed a greater volume of formula and re-
ceived a higher number of feedings per day
may be reflective of a protective response
from historical times when famine was com-
monplace. A classic report of infants un-
der conditions of drought and starvation in
East Africa found that the group of tempera-
mentally easy infants perished whereas most
of the difficult ones survived.35
This same
finding has been replicated in studies ex-
ploring infant behavior and caregiver–infant
interactions from 3 to 6 months of age.36-38
Currently, in developed countries and coun-
tries in transition, excessive infant intake and
weight gain are leading public health con-
cerns for the development of obesity.
Temperament does not directly affect ac-
tivity and metabolism but may be predictive
of behavior patterns or parental care patterns,
both of which may influence energy balance.
Temperament traits are thought to influence
physical health in the areas of accidents and
child abuse, functional abdominal pain, colic,
sleep problems, enuresis, and in nutrition and
growth disorders (failure to thrive, obesity,
and death in famine conditions).27
At a very
basic level, maternal reaction to her infant’s
difficult temperament with additional infant
feeding is very likely the parent’s coping re-
sponse and belief that her infant is distressed
because of hunger.39
Although there was no relationship found
between weight gain and temperament traits,
the lack of change in the mean frequency
of daily bottle-feedings from 4 weeks to 8
weeks to 12 weeks while mean daily total
volume intake did increase was not typical
findings. Typically, the frequency of bottle-
feedings decreases in number, and the total
volume of intake at feedings increases with
infant age. Frequency of feeding decreases be-
cause the growing infant’s stomach capacity
can hold larger volumes at each feeding. Cul-
tural influences also affect infant feeding. Fil-
ipino mothers who perceive their infants to
be “thin” will feed their infant every hour to
promote weight gain.40
This perception origi-
nates from periods of starvation and hunger
in previous generations in the Philippines.
However, it has become increasingly evident
through a growing body of research that faster
early infant weight gain is significantly associ-
ated with later obesity.20,27
Four temperament traits were identified to
be significantly related to infant daily intake
and intake frequency. These were mood, man-
ageability, distractibility, and adaptability. The
data reported here may suggest that some
mothers are responding to their infants’ tem-
perament trait by feeding when the infant is
fussy indicated by the relationship of tem-
perament to feeding frequency. This mater-
nal response to infant temperament ignores
and overrides the infants’ internal cues of
hunger and satiety. This maternal response
to infant temperament has far reaching pub-
lic health implications leading to distorted
eating and child obesity. When parents do
not pay attention to their infants’ internal
cues of hunger and satiety, infants can be

overfed and even force-fed. The longer term
effects may cause disinhibited eating, eat-
ing in the absence of hunger, and obesity.
Anticipatory guidance to those parents with
infants possessing these traits may help to pre-
vent overfeeding and/or the start of harmful
conditioned responses.41
Two infant temperament studies have
shown variation in how infant temperament
affects parenting.42,43
In a literature review,
parents reported that information provided
through anticipatory guidance about tem-
perament influenced their parenting skills.44
Studies suggest that parents should (1)
recognize that infant distress is a signal to
act, (2) respond quickly as waiting does not
reduce crying, and (3) use strategies such as
rocking, presenting alternative activities, and
providing a pacifier, to reduce the crying and
fussing.44
Early identification of early behav-
ioral risk factors such as infant temperament
may set the stage for preventive efforts and
possible treatment to deter the excess intake
that leads to excess weight gain.41,45
Results from this study emphasize the
need in public health and clinical practice
(nutrition, pediatrics) for health profession-
als to offer anticipatory feeding guidance
strategies and counseling to new parents.
These strategies will encompass the earliest
of obesity prevention intervention beginning
in infancy. Anticipatory guidance has appli-
cation in many health care settings such as
primary care practice, hospitals, well-baby
clinics, emergency departments, community
public health clinics and women, infants,
and children clinics, parenting classes, and
prenatal classes to name a few.
A study limitation included the lack of sig-
nificant correlation between infant weight
gain and temperament trait, which may have
been due to the lack of sensitivity of weight
and height measurements and/or the target
period to measure excess weight gain rate was
not accurate (not within the first 3 months of
life). Other investigators have found a correla-
tion between infant caloric intake at 4 months
of age and later development of obesity.14
Instead of measuring weight, length, and head
circumference, body composition measure-
ments (fat mass and fat-free mass) may be a
more sensitive indicator of excess adipose tis-
sue accumulation. In addition, additional con-
founding variables were not controlled that
may affect infant temperament such as genet-
ics and environment. Another study limitation
involved the population ethnicity. It was ho-
mogeneous for the Filipino race, which limits
generalizability to other infant populations.
Given that all of the study infants were fed
infant formula, the results may only be gen-
eralized to formula-fed infants. The Interna-
tional Labor Organization showed that exclu-
sive breastfeeding rates rose from 36% in 2008
to 47% in 2011.
CONCLUSION
This is the first study to demonstrate an
association between temperament and fre-
quency and amount of infant formula feeding.
The results of this study did not demonstrate
a relationship between temperament trait and
infant weight gain. The 4 temperament traits
identified to be significantly related to infant
intake and intake frequency were (1) mood,
(2) manageability, (3) distractibility, and
(4) adaptability. The results of this study
suggest that some mothers respond to their
infants’ temperament trait by feeding or
overfeeding when their infant is fussy, shown
by the relationship of temperament to feeding
frequency. Although a correlational analysis
does not demonstrate cause and effect, there
is a place for these studies. Multivariate
obesity predicting tools valid for newborns
offer optimization of results reliability and ro-
bustness by the adoption of several methods
of high-impact prediction of early obesity.
These tools and use of traditional, routinely
available risk factors should guide future
efforts toward randomized trials of very early
preventive approaches for identifying high-
risk individuals to help combat the obesity
epidemic.46

REFERENCES
1. World Health Organization. Global Increase in
Prevalence of Worldwide Obesity Press Release.
WHO: Geneva; 2004.
2. International Obesity Task Force Prevalence
Data. Global childhood overweight. http://www
.oecd.org/. Published 2011. Accessed August 31,
2013.
3. Ogden CL, Carroll MD, Curtin LR, Lamb MM,
Flegal KM. Prevalence of high body mass index
in US children and adolescents, 2007–2008. JAMA.
2010;303(3):242-249. doi:10.1001/jama.2009.2012.
4. Davison KK, Birch LL. Childhood overweight:
a contextual model and recommendations
for future research. Obes Rev. 2001;2:159-171.
doi:10.1046/j.1467-789x.2001.00036.x.
5. Agras WS, Mascola AJ. Risk factors for childhood
overweight. Curr Opin Pediatr. 2005;17:648-652.
doi:10.1097/01.mop.0000172818.87261.ab.
6. Agras WS, Hammer LD, McNicholas F, Kraemer HC.
Risk factors for childhood overweight: a prospective
study from birth to 9.5 years. J Pediatr. 2004;145:20-
25. doi:10.1016/j.jpeds.2004.03.023.
7. Botton J, Heude J, Maccario J, Ducimetière P, Charles
MA, and the FLVS Study group. Postnatal weight and
height growth velocities at different ages between
birth and 5 y and body composition in adolescent
boys and girls. Am J Clin Nutr. 2008;87:1760-1768.
8. Eriksson J, Forsen T, Osmond C, Barker D. Obesity
from cradle to grave. Int J Obes Relat Metab Disord.
2003;27:722-727. doi:10.1038/sj.ijo.0802278.
9. Ekelund U, Ong K, Linne Y. Both infancy and child-
hood weight gain predict obesity risk at age 17 years:
prospective birth cohort study (SWEDES). Obes Res.
2004;12:A186.
10. Stettler N, Kumanyika SK, Katz SH, Zemel BS,
Stallings VA. Rapid weight gain during infancy and
obesity in young adulthood in a cohort of African
Americans. Am J Clin Nutr. 2003;77:1374-1378.
11. Stettler N, Zemel BS, Kumanyika S, Stallings VA. In-
fant weight gain and childhood overweight status in a
multicenter, cohort study. Pediatrics. 2002;109:194-
199. doi:10.1542/peds.109.2.194.
12. Stettler N, Stallings VA, Troxel AB, et al. Weight
gain in the first week of life and overweight
in adulthood. Circulation. 2005;111:1897-1903.
doi:10.1161/01.CIR.0000161797.67671.A7.
13. Stettler N, Tershakovec AM, Zemel BS, et al. Early
risk factors for increased adiposity: a cohort study
of African American subjects followed from birth to
young adulthood. Am J Clin Nutr. 2000;72:378-383.
14. Ong KK, Emmett PM, Noble S, Ness A, Dunger
DB; ALSPAC Study Team. Dietary energy intake
at the age of 4 months predicts postnatal weight
gain and childhood body mass index. Pediatrics.
2006;117(3):e503-e508. doi:10.1542/peds.2005-
1668.
15. Mei Z, Grummer-Strawn LM, Scanlon KS. Does over-
weight in infancy persist through the preschool
years? An analysis of CDC Pediatric Nutrition Surveil-
lance System data. Soz Praventivmed. 2003;48:161-
167. doi:10.1007/s00038-003-2022-x.
16. Toschke AM, Grote V, Koletzko B, von Kries R.
Identifying children at high risk for overweight
at school entry by weight gain during the first 2
years. Arch Pediatr Adolesc Med. 2004;158:449-452.
doi:10.1001/archpedi.158.5.449.
17. Baird J, Fisher D, Lucas P, Kleijnen J, Roberts
H, Law C. Being big or growing fast: systematic
review of size and growth in infancy and later
obesity. BMJ. 2005;331(7522):929. doi:10.1136/
bmj.38586.411273.E0.
18. Reilly JJ, Armstrong J, Dorosty AR, et al.
Early life risk factors for obesity in child-
hood: cohort study. BMJ. 2005;330(7504):1357.
doi:10.1136/bmj.38470.670903.E0.
19. Lucas A, St., James-Roberts I. Crying, fussing, and
colic behaviour in breast- and bottle-fed infants. Early
Hum Dev. 1998;53(1):9-18.
20. Barr RG, Kramer MS, Pless IB, Boisjoly C, Leduc
D. Feeding and temperament as determinants of
early infant crying/fussing behavior. Pediatrics.
1989;84:514-521.
21. Carey WB. Temperament and increased weight gain
in infants. J Dev Behav Pediatr. 1985;6(3):128-131.
doi:10.1097/00004703-198506000-00006.
22. Capeding R, Gepanayao CP, Calimon N, et al. Lutein-
fortified infant formula fed to healthy term infants:
evaluation of growth effects and safety. Nutr J.
2010;9:22-31. doi:10.1186/1475-2891-9-22.
23. Medoff-Cooper B, Carey WB, McDevitt SC. The early
infancy temperament questionnaire. J Dev Behav
Pediatr. 1993;14(4):230-235.
24. Wiesenfeld AR, Malatesta CZ, Whitman PB, Granrose
C, Uili R. Psychophysiological response of breast- and
bottle-feeding mothers to their infants’ signals. Psy-
chophysiology. 1985;22:79-86. doi:10.1111/j.1469-
8986.1985.tb01563.x.
25. Vandiver TA. Relationship of mother’s percep-
tions and behaviors to the duration of breastfeed-
ing. Psychol Rep. 1997;80:1375-1384. doi:10.2466/
pr0.1997.80.3c.1375.
26. Britton JR, Britton HL, Gronwaldt V. Breast-
feeding, sensitivity and attachment. Pediatrics.
2006;11:8e1436.
27. Darlington AS, Wright CM. The influence of temper-
ament on weight gain in early infancy. J Dev Behav
Pediatr. 2006;27(4):329-335. doi:10.1097/000047
03-200608000-00008.

28. Carey WB. Clinical use of temperament data in pedi-
atrics. J Dev Behav Pediatr. 1985;6(3):137-142.
29. Carey WB. Validity of parental assessments of devel-
opment and behavior. Am J Dis Child. 1982;136:
97-99.
30. Centers for Disease Control and Prevention. SAS
program for the CDC growth charts. http://
www.cdc.gov/nccdphp/dnpa/growthcharts/
resources/sas.htm. Accessed August 31, 2013.
31. World Health Organization. Growth standards.
http://www.who.int/childgrowth/en/.
32. Carey WB, Hegvik RL, McDevitt SC. Tempera-
mental factors associated with rapid weight gain
and obesity in middle childhood. J Dev Behav
Pediatr. 1988;9(4):194-198. doi:10.1097/00004703-
198808000-00003.
33. Wells JC, Stanley M, Laidlaw AS, Day JM, Stafford
M, Davies PS. Investigation of the relationship be-
tween infant temperament and later body composi-
tion. Int J Obes Relat Metab Disord. 1997;21(5):400-
406. doi:10.1038/sj.ijo.0800420.
34. Ravaja N, Keltikakangas-Jarvinen L. Temperament
and metabolic syndrome precursors in children: a
three year follow-up. Prev Med. 1995;24:518-527.
doi:10.1006/pmed.1995.1082.
35. deVries MW. Temperament and infant mortality
among the Masai of East Africa. Am J Psychiatry.
1984;141:1189-1194.
36. Singhal A, Kennedy K, Lanigan J, et al. Nutrition
in infancy and long-term risk of obesity: evidence
from 2 randomized controlled trials. Am J Clin
Nutr. 2010;92:1133-1144. doi:10.3945/ajcn.2010
.29302.
37. Rahmanifar A, Kirksey A, Wachs TD, et al. Diet dur-
ing lactation associated with infant behavior and
caregiver-infant interaction in a semi-rural Egyptian
village. J Nutr. 1993;123:164-175.
38. Slining MM, Adair L, Goldman BD, Borja J, Bentley
M. Infant temperament contributes to early infant
growth: a prospective cohort of African American
infants. Int J Behav Nutr Phys Act. 2009;6:51-61.
doi:10.1186/1479-5868-6-51.
39. Cameron JR, Rice DC. Developing anticipatory
guidance programs based on early assessment
of infant temperament: two tests of a preven-
tion model. J Pediatr Psychol. 1986;11:221-234.
doi:10.1093/jpepsy/11.2.221.
40. Holub SC, Dolan EA. Mothers’ beliefs about infant
size: associations with attitudes and infant feeding
practices. J Appl Dev Psych. 2012;33:158-164.
41. Paul IM, Bartok CJ, Downs DS, Stifter CA, Ventura AK,
Birch LL. Opportunities for the primary prevention
of obesity during infancy. Adv Pediatr. 2009;56:107-
133. doi:10.1016/j.yapd.2009.08.012.
42. Wasser H, Bently M, Bonja J, et al. Infants perceived
as “fussy” are more likely to receive complementary
foods before 4 months. Pediatrics. 2011;127:229-
237. doi:10.1542/peds.2010-0166.
43. Redsell SA, Atkinson P, Dilip N, Siriwardena AN, Swift
JA, Glazebrook C. Parent’s beliefs about appropriate
infant size, growth and feeding behavior: implica-
tions for the prevention of childhood obesity. BMC
Public Health. 2010;10:711-721. doi:10.1186/1471-
2458-10-711.
44. Regalado M, Halfon N. Primary care services promot-
ing optimal child development from birth to 3 years:
review of the literature. Arch Pediatr Adolesc Med.
2001;155:1311-1322.
45. Dattilo AM, Birch L, Krebs NF, Lake A, Taveras EM,
Saavedra JM. Need for early interventions in the
prevention of pediatric overweight: a review and
upcoming directions. J Obes. 2012;2012: 123023.
doi:10.1155/2012/123023.
46. Morandi1 A, Meyre D, Lobbens S, et al. Es-
timation of newborn risk for child or ado-
lescent obesity: lessons from longitudinal
birth cohorts. PLoS ONE. 2012;7(11):e49919.
doi:10.1371/journal.pone.0049919.

DavisEarly_Infant_Temperament2013TICN

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to DavisEarly_Infant_Temperament2013TICN

Similar to DavisEarly_Infant_Temperament2013TICN (20)

DavisEarly_Infant_Temperament2013TICN