Original
Article
. . . . . . . . . . . . . .
Comparison of Breast- and Formula-Fed Normal Newborns
in Time to First Stool and Urine
Mirela Metaj, MD
Nirupama Laroia, MD
Ruth A. Lawrence, MD
Rita M. Ryan, MD
BACKGROUND:
Although the time to an infant’s first stool is used as a marker for a
normal GI tract, there have been no studies to date which have evaluated
the contribution of the first 24 hours’ feeding type to time to first
meconium stool, directly comparing breast feeding vs formula feeding.
OBJECTIVE:
To compare breast- and formula-fed healthy infants Z34 weeks gestation
in time to first stool and urine.
STUDY DESIGN:
A chart review of 1000 consecutive infants Z34 weeks of gestational age
admitted to the normal newborn nursery of Children’s Hospital of Buffalo
from June to October 2000. Infants (n ¼ 979) were grouped based on
feeding type in the first 24 hours: breast-fed (n ¼ 211), formula-fed
(n ¼ 540), and mixed feeding (n ¼ 228); n ¼ 21 excluded for Neonatal
Intensive Care Unit admission. We initially compared the time to first
stool and urine between the breast- and formula-fed groups and then
examined multiple maternal and infant demographic and clinical factors
for their effect on time to first stool using univariate and multivariate
analyses.
RESULTS:
Breast-fed infants were fed earlier and more frequently than formula-fed
but there was no significant difference in time to first stool (7.6 vs 7.9
hours). Breast-fed infants were earlier in time to first urine (p ¼ 0.03)
(7.3 vs 8.5 hours). In multiple regression analysis, gestational age was the
only significant (p ¼ 0.000) factor in predicting time to first stool.
Department of Pediatrics (M.M., R.M.R.), State University of New York at Buffalo and Children’s
Hospital of Buffalo, Buffalo, NY, USA; University of Rochester (N.L., R.A.L.), Rochester, NY, USA;
and Visiting Fulbright Scholar (M.M.), Obstetric Hospital, Fier, Albania.
Address correspondence and reprint requests to Rita M. Ryan, MD, Associate Professor of
Pediatrics (Neonatology), Department of Pediatrics, State University of New York at Buffalo,
Children’s Hospital of Buffalo, 219 Bryant Street, Buffalo, NY 14222, USA
624
CONCLUSION:
Type of feeding did not predict time to first stool but gestational age was
important even in this near-term and term population of infants Z34
weeks gestation.
Journal of Perinatology (2003) 23, 624–628. doi:10.1038/sj.jp.7210997
INTRODUCTION
Passage of an infant’s first meconium stool is often used as a
screen for a normal GI tract.1–4 Two studies in the American
population state that 94%5 and 98.5%6 of term infants had their
first meconium stool by 24 hours. These studies were performed
some time ago when practices in the normal newborn nursery were
likely quite different from current practices. These studies did not
examine the type and schedule of feedings. In recent studies in
Asia,2–4 the contribution of different physiologic factors (sex,
gestational age, birth weight, type of delivery, Apgar scores) to time
to first meconium stool was evaluated. There have been no studies
that have evaluated the contribution of first 24 hours’ feeding type
to time to first meconium stool, directly comparing breastfeeding vs
formula feeding. With an increase in the last two decades in
breastfeeding in the US, it is important to document the effect, if
any, on the time to first stool and first urine since the content,
frequency and volume of feeds are different in breast- and formula-
fed infants. It is possible that breast-fed infants may have a delay in
time to first stool because there is a lower volume of feeding in the
first 24 hours. Alternatively, breast-fed infants are often fed earlier
and more frequently, and colostrum is considered by some to be a
cathartic, so breast-fed infants may have a shorter time to first
stool. Similarly, the type of feeding may affect time to first urine if
breast-fed infants receive lower volume of feeding in the first 24
hours. The purpose of our study was to compare breast- and
formula-fed healthy infants Z34 weeks of gestational age in time
to first stool and first urine.
METHODS
Most infants of 34 weeks gestation and older are cared for in the
newborn nursery. We reviewed retrospectively all maternal and
infant charts of 1000 consecutive infants Z34 weeks gestational
age admitted and discharged from the normal newborn nursery of
Journal of Perinatology 2003; 23:624–628
r 2003 Nature Publishing Group All rights reserved. 0743-8346/03 $25
www.nature.com/jp
Comparison of Breast- and Formula-Fed Newborns
Children’s Hospital of Buffalo (June to October 2000). Infants were
excluded if they were transferred to the Neonatal Intensive Care
Unit (NICU) at any time, even if they were ultimately discharged
from normal newborn nursery. Infants were defined in three
groups based on feeding type during the first 24 hours of age:
breast-fed, formula-fed, and mixed feeding. The infants in the
breast-fed group were fed exclusively with breast milk for the first
24 hours (although 25% had dextrose water as their first feeding).
The infants in the formula-fed group were fed exclusively with
formula for the first 24 hours (although 98% had dextrose water as
their first feeding). The infants in the mixed feeding group
received both breast milk and formula during the first 24 hours
(and 60% had dextrose water, while the remainder had breast milk
as their first feeding).
Data collection. Maternal demographic variables collected
included maternal age, race, gravity, parity, illicit drug
use, intrapartum drug use (narcotics, antibiotics), mode of
delivery, meconium stained amniotic fluid, and maternal diabetes.
Infant variables collected included gender, Apgar scores, gestational
age, birth weight, congenital malformations, age at first feed,
number of feeds from 0 to 12 hours of age, 12 to 24 hours
of age, 24 to 48 hours of age, age at discharge, age at first
and second urine, age at first and second stool, description of first
stool; meconium staining of the amniotic fluid was not considered
a first stool.
Statistics. Data entry and analysis were performed using STATA
(State College, TX). Variables of exclusively breast-fed and
exclusively formula-fed infants were compared using t-test, linear
regression and w2 statistic as appropriate with p<0.05 as the
significance level. A p-value of 0.000 reported by STATA is
equivalent to a p-value <0.00005. All factors were examined by
univariate analysis in the total population for the effect on time to
first stool and first urine. Variables with a significance level r0.1
on univariate analysis were then examined in a multivariate
analysis using stepwise linear regression to evaluate independent
predictors of time to first stool or first urine, with a final
significance level of p<0.05.
RESULTS
Charts for 1000 consecutive infants were reviewed. We excluded 21
infants who were admitted to the NICU, then returned to newborn
nursery prior to discharge. Of the remaining 979 infants, 211 were
in the breast-fed group, 540 were in the formula-fed group, and
228 were in the mixed feeding group. In our population there
were 496 boys and 483 girls. In all, 723 infants (74%) were born
vaginally, while 256 (26%) were born by Cesarean section. A total
of 55 infants (6%) were born to diabetic mothers (47 with
gestational diabetes). There was meconium staining of amniotic
fluid in 148 (15%) cases. A total of 31 infants (3%) had congenital
anomalies. There was one case each of Trisomy 21 and cleft palate,
and six cases of hypospadias. Intrapartum nalbuphine
Journal of Perinatology 2003; 23:624–628
Metaj et al.
hydrochloride (nubain) was received by 269 (27%) of mothers and
231 (24%) received antibiotics prior to delivery. In total, 97%
of our infants had their first stool by 24 hours and 99.8% of
them had stool by 36 hours. A total of 74% of our infants had
passed urine by 12 hours of age and 97.9% of them by 24 hours.
Figure 1a and b provides more detailed information.
Comparison of Breast- and Formula-fed Infants
Important maternal and infant demographic and medical
information is reported for the exclusively breast-fed and
exclusively formula-fed group (first 24 hours) in Table 1. Stooling
and urination results by type of feeding are reported in Table 2.
Breast-fed infants were initially fed significantly earlier than
formula-fed infants. They also had an average of one more feeding
in the first 12 and second 12 hours of life compared to formula-fed
infants. Breast-fed infants were discharged earlier. The breast-fed
infants lost more weight than the formula-fed infants and were
96% of birthweight vs 98% of birthweight, respectively, at the time
a
120
Cumulative percentage of infants
100
80
60
40
20
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
Hours of age
b
120
Cumulative percentage of infants
100
80
60
40
20
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
Hours of age
Figure 1. These graphs depict the cumulative percentage of infants
who had their first stool (a) or first urine (b) by a certain hour of age.
For graph a, 99.7 infants had their first stool by 34 hours; there were an
additional three infants who had their first stool later (36, 37, and 61
hours). For graph (b), 99.8 infants had their first urine by 34 hours;
there were an additional two infants who had their first stool later (35
and 37 hours).
625
 Metaj et al. Comparison of Breast- and Formula-Fed Newborns

Table 1 Characteristics of breast-fed and formula-fed infants

Factors which Contribute to Time to First Stool
After the direct comparison of breast- and formula-fed infants was
Breast fed Formula fed performed, all 979 infants were studied by univariate analysis for
(n=211) (n=540) other factors that may contribute to time to first stool. p-Values for
Maternal age (years, mean, SD)* 29.1±6.7 26.2±6.5 each variable are represented in Table 3. As the presence of
Maternal race* meconium in the amniotic fluid would appear to be an indicator of
Caucasian 136 (64%) 186 (34%) gut peristalsis (as well as fetal stress), it is interesting to note that
African-American 53 (25%) 269 (50%) infants who passed meconium in utero passed stool significantly
Hispanic 11 (5%) 61 (11%) earlier than non-meconium-stained infants, as described in Table 3.
Illicit drug use* 8 (3.8%) 64 (11.9%) We then performed multivariate analysis including all the variables
Intrapartum narcotics 62 (29.3%) 145 (26.9%)
with a significant level r0.1 in univariate analysis. In the
Intrapartum antibiotics 42 (19.9%) 117 (21.7%)
Cesarean section delivery 51 (24.2%) 139 (25.7%)
Meconium-stained amniotic fluid 33 (15.6%) 80 (14.8%)
Maternal diabetes 8 (3.8%) 24 (4.4%) Table 3 Univariate Analysis of Maternal and Infant Factors
Gender Predicting Time to First Stool
Male 102 (48%) 268 (50%)
Female 109 (52%) 272 (50%) Factor p-Value Time to % of infants
Gestational age (weeks, mean, SD)* 39.4±1.2 weeks 38.9±1.6 weeks first stool >17 hours
<37 weeks* 7 (<3%) 59 (11%) (h, mean, SD) at first stool
Birth weight (g, mean, SD)* 3416±458 3218±555 (E90th percentile)
<2500 g* 3 (1.5%) 59 (11%) Maternal age* 0.08
Congenital malformations 6 (2.8%) 20 (3.7%) Racew 0.25
*p<0.05. Illicit drug use 0.77
Intrapartum narcotics 0.46
Intrapartum antibiotics 0.92
Mode of delivery* 0.03
Table 2 Comparison of breast-fed and formula-fed infants Vaginal 7.5±6.7 8.7%
Breast fed Formula fed p-Value Cesarean section 8.6±7.7 10.9%, p=0.292
(n=211) (n=540) by w2
Meconium-stained AFz,* 0.000
Age (h) at first feeding* 1.8±1.5 3.7±1.3 0.0000 Yes 3.6±6.1 4.7%
Number of feeds 0–12 hours* 3.4±0.9 2.8±0.7 0.0000 No 8.5±6.8 10.1%, p=0.038
Number of feeds 12–24 hours* 3.6±1.0 3.1±0.7 0.0000 by w2
Number of feeds 24–48 hours* 6.5±1.9 5.7±1.4 0.0000 Maternal diabetes 0.97
Age (h) at discharge* 53.6±18.8 62.7±36.2 0.0005 Gender 0.11
Ratio second weight/birth weight* 0.96±0.02 0.98±0.03 0.0000 Gestational age*y 0.000
Age (h) at first stool 7.6±6.9 7.9±7.2 0.64 Prematurity 0.000
No. (%) had not stooled GA <37 weeks 12.7±7.6 26.7%
By 17 hours (90th%) 18 (8.5%) 52 (9.6) 0.64 GA Z37 weeks 7.3±6.7 7.6%, p=0.000 by w2
By 24 hours (97.5th%) 6 (2.8%) 17 (3.2%) 0.83 Birth weight* 0.000
Age (h) at first urine* 7.3±7.3 8.5±7.2 0.03 Low birth weight 0.000
No. (%) had not urinated <2500 g 10.7±8.8 20.5%
By 18.5 hours (90th%) 17 (8.1%) 54 (10.0) 0.41 Z2500 g 7.5±6.7 8.3%, p=0.000 by w2
By 24 hours (98th) 7 (3.3%) 9 (1.6%) 0.16 Congenital 0.07
*p<0.05. malformations*
Age at first feeding 0.38
Number of feeds 0–12 0.14
hours
of the second weight. The average age (h±SD) of the baby at the Type of feeding first 0.89
time of the second weight was 23.8±7.2 and 23.3±7.9 for breast- 24 hours (ANOVA)
and bottle-fed infants, respectively. There was no difference in time *Used in multiple regression analysis.
w
to first stool between two groups, and no difference in the percent z
Caucasian vs African-American only.
Amniotic fluid.
of infants who stooled by 24 hours of age. Breast-fed infants were y
Remained significant on multiple regression analysis.
significantly earlier in time to first urine (p<0.05).

626 Journal of Perinatology 2003; 23:624–628

Comparison of Breast- and Formula-Fed Newborns Metaj et al.

16
Table 4 Univariate Analysis of Maternal and Infant Factors
14
Predicting Time to First Urine
12
Factor p-Value Time to % of infants >18

Hours of age
10
first urine hours at first
8
(h, mean, SD) urine (E90th
percentile) 6

4
Maternal age 0.14
Race*w 0.02 2

Caucasian 7.5±6.8 0
African-American 8.7±7.9 34 35 36 37 38 39 40 41 42
Gestational Age
Illicit drug use 0.63
Intrapartum narcotics 0.22 Figure 2. This graph depicts the average time to first stool for each
Intrapartum antibiotics 0.50 gestational age. Each age plotted represents all infants at that week of
Mode of delivery*z 0.000 gestational age; for example, 36 weeks represents infants 36 0/7 weeks
Vaginal 8.7±7.3 11.9% to 36 6/7 weeks inclusive. Standard deviations were omitted from the
Cesarean section 6.3±7.0 5.8%, p=0.024 graph but ranged from 3.8 to 9.9 hours.
by w2
Meconium-stained AFy 0.12
Maternal diabetes* 0.055 represented in Table 4. We then performed multivariate analysis
Gender 0.36 including all the variables with a significant level r0.1 in
1 minute Apgar score 0.24 univariate analysis. In the multiple regression analysis, three
5 minute Apgar score 0.40 factors remained significant in time to first urine: gestational age
Gestational age*z 0.000 (p ¼ 0.000), mode of delivery (p ¼ 0.000), and type of feeding in
Prematurity 0.001
first 24 hours (p ¼ 0.004).
GA <37 weeks 5.5±6.4 4.7%
GA Z37 weeks 8.3±7.3 9.9%, p=0.114
by Fisher’s Gestational age
Birth weight* 0.001 Since the gestational age was highly significant in predicting time
Low birth weight 0.001 to first stool, we compared preterm and term infants in this
<2500 g 5.4±6.5 5.1% population by type of feeding. In term infants, there was no
Z2500 g 8.3±7.3 9.8%, p=0.178 difference in mean (SD) time to first stool which was 7.2 (6.4)
by Fisher’s hours in breast-fed infants (n ¼ 204) and 7.2 (6.9) hours in
Congenital malformations 0.14
formula-fed infants (n ¼ 481). However, there was a statistical
Age at first feeding* 0.06
difference in time to first urine which was 7.2 (7.3) hours in
Number of feeds 0–12 0.01
hours*
breast-fed infants vs 8.7 (7.1) hours in formula-fed infants
Type of feeding first 24 0.09 (p<0.05). In preterm infants, there was no difference in time to
hours*z (ANOVA) first urine which was 6.3 (5.8) hours in breast-fed infants (n ¼ 7)
Breast fed (n=211) 7.2±7.3 8.5% and 6.4 (6.7) hours in formula-fed infants (n ¼ 59). Although
Breast and formula 7.8±7.6 9.2% there was a trend in the breast-fed premature infants to have a
(n=228) small delay in time to first stool (16.1 (13) hours vs 12.5 (7.5) in
Formula fed (n=540) 8.5±7.1 10%, p=0.71 by w2 formula-fed infants), since most premature infants were formula-
*Used in multiple regression analysis. fed the n is small and no conclusions can be drawn. The
w
Caucasian vs African-American only. importance of gestational age in time to first stool in the total
z
Remained significant on multiple regression analysis. population is demonstrated by Figure 2.

multiple regression analysis, only gestational age remained

significant (p ¼ 0.000) in predicting time to first stool.
Factors which Contribute to Time to First Urine
A similar analysis was carried out for time to first urine examining
all 979 infants by univariate analysis for factors that may
contribute to time to first stool. p-Values for each variable are
DISCUSSION
Time to first stool is not related to the age at first feeding or to type
of feeding. In total, 33% of our infants had their first stool prior to
any feeding, and despite breast-fed infants feeding twice as early as
formula-fed infants, there was no difference in time to first stool
between the two groups. Therefore, factors other than timing and

Journal of Perinatology 2003; 23:624–628 627

 Metaj et al.
type of first feed must be important in triggering gut motility in the
healthy newborn infant.
Infants with meconium-stained fluid stooled significantly
sooner than infants with no meconium at birth. There was no
difference in time to first stool between vaginally and cesarean
section born infants as has been reported previously.4
Motor activity is detectable in the small intestine as early as 26
weeks. These random contractions progress to more regular bursts
of motor activity by 30 weeks. These are replaced by migrating
motor complexes by 33 weeks gestation. Coordinated sucking and
swallowing also required for independent feeding is not achieved
until 34 to 35 weeks gestation. This gestational age seems to
coincide with a significant increase in defecation rate, a surge in
concentration of circulating intestinal regulatory polypeptides in
response to feeding.7 Although sometimes clinicians associate
stooling to feeding, in our study one-third of the infants had
stooled prior to any feeding. Feeding may be a more important
factor in subsequent stooling.
A total of 10% of our population was premature (34 to 36 6/7
weeks). Overall, premature infants stooled significantly later than
full-term infants did (p ¼ 0.000), implying that the maturation of
the GI tract (especially small gut and colon) continues until late in
pregnancy. Only a small number of these premature infants were
breast fed; therefore, no conclusions could be drawn regarding type
of feeding. Kumar et al.8 have previously suggested that delayed
passage of stools is both a function of illness severity and
gestational immaturity. Our study shows that this is a valid
628
Comparison of Breast- and Formula-Fed Newborns
conclusion even in otherwise healthy preterm infants cared for in
the normal newborn nursery. Overall, breast-fed infants passed
urine sooner than formula-fed infants; most of these infants were
term. No conclusion can be drawn regarding type of feeding and
passage of urine in the premature infants. We conclude that despite
differences in timing and number of feedings, breast- and formula-
fed term infants have no difference in time to first stool, but
gestational age remains important even in this near-term and term
healthy population (Z34 weeks).
References
1. Taeusch HW, Ballard RA. Avery’s diseases of newborn. WB Saunders
Company; Philadelphia, PA 1998. p. 902–4.
2. Chih TW, Teng RJ, Wang CS, Tsou Yau KI. Time of the first urine and the
first stool in Chinese newborns. Acta Ped Sin 1991;32:17–23.
3. Tejajev A, Siripoonya P. The times of passage of the first urine and the first
stool by Thai newborn infants. J Med Ass Thailand 1984;67(2):86–8.
4. Tateishi H, Yamauchi Y, Yamanouchi I, Khshaba MT. Effect of mode of
delivery, parity and umbilical blood gas on first meconium passage in full-
term healthy infants. Biol Neonate 1994;66:146–9.
5. Sherry SN, Kramer I. The time of passage of the first stool and first urine by
the newborn infant. J Pediatr 1955;46:158–9.
6. Clark DA. Times of first void and first stool in 500 newborns. Pediatrics
1977; 60(4):457–9.
7. Weaver LT, Lucas A. Development of bowel habit in preterm infants. Arch
Dis Child 1993;68:317–20.
8. Kumar SL, Dhanireddy R. Time to first stool in premature infants: effect of
gestational age and illness severity. J Pediatr 1995;127:971–4.
Journal of Perinatology 2003; 23:624–628