Determinants of Milk Flow
Role of Hole Size and Nipple Thickness
Oommen P. Mathew, MD
\s=b\ The aim of the present study was to
elucidate the role of hole size and thick-
ness in determining milk flow through nip-
ple units during bottle feeding. Commonly
used standard nipple units (SMA single\x=req-\
hole, Enfamil single-hole, and Twist-on)
for term and preterm infants, as well as
Nuk-type nipple units (SMA Nuk, Enfamil
Natural, and Nuk) were tested. The size of
the nipple hole and wall thickness were
determined for each nipple unit. Airflow
was measured by forcing pressurized air
through the feed hole. Simulated sucks
were used to measure the milk flow. A
marked variability in airflow and milk flow
was observed within and among the vari-
ous types of nipple units studied. Within
each type of nipple unit, both milk flow and
airflow measurements correlated well
with hole size. The thickness of the nipple
units contributed minimally to the ob-
served variability. We conclude that differ-
ences in hole size primarily account for
the observed variability in milk flow. This
finding may be clinically important in that
rapid milk flow can lead to apnea and bra-
dycardia in some preterm infants. The
above observations imply that design
changes are necessary to reduce the vari-
ability of milk flow within each nipple type.
Moreover, milk-flow measurements made
using a simple mechanical system and
airflow measurements used by the indus-
try are equally sensitive to evaluate nipple
flow.
(AJDC. 1990;144:222-224)
number of different types of nipple
units are available for use in neo¬
nates; some are specially designed for
premature infants, whereas others are
designed to resemble the human nipple.
At present, most preterm infants are
Accepted for publication September 18,1989.
From the Department of Pediatrics, University
of Texas Medical Branch, Galveston.
Reprint reqests to Perinatal Pediatrics, Univer-
sity of Texas Medical Branch, Galveston, TX 77550
(Dr Mathew).
Downloaded From: http://archpedi.jamanetwork.com/ by a Missouri S & T User on 06/22/2015
Through Nipple
fed with nipple units designed for pre¬
term infants. However, the functional
superiority of these types of nipples
over other types for these infants has
not been adequately documented. For
instance, the assumption has been made
that preterm infants generate less pres¬
sure per suck and the preterm nipple
units are therefore designed to deliver
higher flow compared with nipple units
designed for term infants. However, it
is not clear whether there is any advan¬
tage in increasing the milk flow in pre¬
term infants. On the contrary, higher
milk flow may produce some adverse
effects by altering the breathing pat¬
tern significantly.
The milk flow through nipple units
can be altered by a number of factors:
the characteristics of the nipple, the ri¬
gidity of the container, and the sucking
pressure generated by the infant.1"3
Since the pressure generated by the in¬
fant varies from suck to suck and from
infant to infant, we have developed a
system with which we can evaluate the
different nipple units at a constant suck¬
ing frequency and pressure. ' Since rigid
containers are used, the differences in
milk flow can be attributed solely to
differences in nipple characteristics.
Recently, we reported a marked vari¬
ability in milk flow within the same type
and among the different types of nipple
units. ' We speculated that the size of the
nipple hole is the most important vari¬
able in determining milk flow; the thick¬
ness of the nipple unit may also contrib¬
ute to variability.
The aim of the present study was to
establish the role of these two variables
in determining milk flow. An additional
objective of the study was to compare
the milk-flow determinations of the sim¬
ple mechanical system with airflow
measurements, the industry standard.
Units
MATERIALS AND METHODS
Nipple units commonly used in the feeding
of newborn infants formed the basis of this
study. These nipple units were manufac¬
tured and/or distributed by Mead-Johnson
(Evansville, Ind), Ross Laboratories (Co¬
lumbus, Ohio), and Wyeth (Philadelphia,
Pa), and consisted of standard nipple units
for term and preterm infants as well as the
Nuk-type units. The general characteristics
of these nipple units have been reported
previously.1
The following tests and measurements
were conducted on each nipple: the size of the
nipple hole, the thickness at the tip of the
nipple, the airflow, and the milk flow. The
size of the nipple hole was calculated using a
profile projector (Mitutoyo Manufacturing
Co). The diameter of the hole was measured
along the and y axes. The average of the
two values obtained was taken as the diame¬
ter of the feed hole. The wall thickness was
determined using a Magna-Mike thickness
gauge (Automatic Inspection Device, Ine).
The wall thickness was measured by using a
small steel ball on the inside of the nipple and
a magnetic probe on the outside of the nip¬
ple's bulb. The gauge senses the magnetic
field strength between the ball and the tip of
the probe. Since the magnetic field varies in a
known manner with the distance between
the ball and probe, the wall thickness can be
calculated accurately with a microprocessor.
The airflow was measured by inserting the
nipple into an orifice and then inserting a
plunger into the nipple forcing pressurized
air (42 psi) through the feed hole. The flow
rate of the air passing through the nipple was
measured with an airflow meter.
The milk flow through the nipple was mea¬
sured as reported previously.1 Briefly, the
nipple to be tested was firmly secured on a
bottle containing ready-to-feed formula (En¬
famil, 281 kJ/dL), inverted, and placed on a
rigid container. A negative pressure pulse of
-120 cm H20 at 40 cycles per minute was
then applied to the rigid container with a
breast pump. The number of sucks required
to empty a bottle with 120 mL of formula was
determined and was inversely proportional
 to milk flow. Twenty nipple units were evalu¬
ated in each category. Comparison of Various Types of Nipple Units Used in the Neonatal Period
Stepwise linear regression analysis was Mean +
SD
used to correlate the hole size and wall thick¬
ness to airflow and milk flow rates. A value of Hole Size, Thickness, Airflow,
P<.05 was considered significant. Nipple Units cm cm m Us Milk Flow*
Term Infants
RESULTS Standard
Nipple Units for Term Infants SMA single-hole 0.043 ±0.003 0.077 ±0.012 30.7 + 2.8 777 ±106
Enfamil single-hole 0.036 ±0.004 0.084 ±0.011 24.4 ±3.6 1068 ±185
The results obtained with various Twist-on 0.031 ±0.008 0.127±0.015 5.1 ±2.7 627±268
nipple units are summarized in the Ta¬ Nuk-type
ble. In general, Nuk-type nipples had SMA 0.069 ±0.004 0.132 ±0.015 52.1 ±3.5 135±11
larger feed holes compared with stan¬ Enfamil Natural 0.040 ±0.003 0.116 ±0.020 25.9 ±3.3 342 ±56
dard nipple units for term infants. Nuk 0.055±0.004 0.110±0.012 42.3±3.5 183±49
Among the nipple units for term infants, Preterm Infants
the largest percentage variation in hole Standard
size was observed with the Twist-on SMA single-hole 0.039 ±0.005 0.131 ±0.015 23.8 ±5.3 827 ±180
nipple unit for term infants, made by Enfamil single-hole 0.054 ±0.005 0.115±0.016 43.1 ±2.9 320 ±37
Ross Laboratories. Among the stan¬ Twist-on 0.032 + 0.007 0.111 ±0.022 28.8±10.3 705 ±480
dard types, the greatest thickness at *
Values represent the number of simulated sucks to empty a bottle with 120 mL of formula.
the tip was observed with the Twist-on
nipple unit, and among the Nuk types,
the greatest thickness was observed
with the SMA unit. The SMA single-
hole unit had the largest percentage
variation in thickness among the stan¬
dard units, and the Enfamil Natural had
the largest percentage variation in
thickness among the Nuk-type nipple
units (Table).
Greater hole size was associated with
higher airflow and milk flow (indicated
by a smaller number of simulated
sucks). Regression analysis revealed a
highly significant (P<.001) correlation
between hole size and airflow as well as
between hole size and milk flow for stan¬
dard nipple units (Fig 1). This correla¬
tion was not significantly improved by
the addition of thickness. As for the Fig 1.—Standard nipple unit for term infants. The circle indicates SMA single-hole unit; the
square, Enfamil single-hole unit; and the triangle, Twist-on unit. Open symbols indicate airflow;
standard nipple unit, a highly signifi¬ solid symbols, milk flow. The higher the number of sucks, the lower the milk flow.
cant (P<.005) correlation existed be¬
tween hole size and airflow as well as
between hole size and milk flow for SMA
Nuk units and Enfamil Natural units,
whereas the correlation was less signifi¬
cant (P<.05) for the Nuk nipple distrib¬
uted through Ross Laboratories (Fig 2).
Thickness significantly improved this
correlation with milk flow only with the
SMA Nuk nipple unit.
Nipple Units for Preterm Infants
These nipple units had larger feed
holes compared with their counterparts
for term infants for Enfamil and smaller
feed holes for SMA nipple units (Table).
Both airflow and milk flow were corre¬
Fig 2.—Nuk-type nipple units. Correlations between airflow and hole size and milk flow and hole
spondingly higher with Enfamil and size are shown. Triangles indicate the Nuk nipple distributed through Ross Laboratory. See Fig
lower with SMA units for premature 1 for definitions of other symbols.

Downloaded From: http://archpedi.jamanetwork.com/ by a Missouri S & T User on 06/22/2015

 infants. Again, regression analysis
showed highly significant (P<.001) cor¬
relations between hole size and airflow
as well as between hole size and milk
flow (Fig 3). No significant correlation
between thickness and airflow or milk
flow was observed with any of the nipple
units.
COMMENT
Using simulated sucks, we recently
reported a great variability in milk flow
within and among the various types of
nipple units available for use in new¬
borns.1 Results of the present study
confirm and extend this observation.
The number of simulated sucks fell
within the 2 SDs reported previously.1
The marked variability observed in
both studies underscores the variability
in milk flow seen with these nipple
units. As a result of the earlier study,
we hypothesized that differences in the
size of the nipple hole probably account¬
ed for most of the observed variability in
milk flow and we suggested that rapid
milk flow may lead to apnea and brady¬
cardia in some preterm infants.1 The
present results clearly document that
milk flow and airflow correlate well with
the size of the nipple hole. This correla¬
tion was observed with all the different
types of nipple units studied. The differ¬
ence in thickness (within the ranges
studied) did not significantly improve
the correlation between hole size and
milk flow, except with the SMA Nuk
nipple.
Our results indicate that airflow and
simulated sucks are both equally good in
evaluating functional characteristics of
the nipple (Figs 1 through 3). However,
simulated sucks employ relatively sim¬
ple methods compared with the airflow
measurements used by the industry.
Moreover, since simulated suck evalu¬
ates milk flow rather than airflow, and
since it employs negative pressure
pulse rather than constant positive
pressure, this form of evaluation resem¬
bles more closely the physiological con¬
ditions observed during bottle feeding.
We did not measure directly the dis-
tensibility of the nipple units. Instead,
we chose to measure the wall thickness
at the tip of the nipple, which is an indi¬
rect measure of distensibility. For most
Downloaded From: http://archpedi.jamanetwork.com/ by a Missouri S & T User on 06/22/2015
Fig 3.—Standard nipple units for preterm infants. See Fig 1 for definitions of symbols.
of the nipple units studied, there was a
poor correlation between nipple thick¬
ness and airflow or milk flow. Hence,
wall thickness at the tip, at least in the
ranges studied, does not appear to be an
important determinant of milk flow.
Even though it is clear that differ¬
ences in hole size are primarily responsi¬
ble for the variation in milk flow, the
various tests used did not reveal the
reason for this variability. One likely
explanation is the methods used in mak¬
ing the feed hole. The feed holes are
drilled mechanically, a method that
would be expected to produce signifi¬
cant variation in hole size. Innovations
in this technique, such as the use of
laser, may help reduce this variation in
hole size. Preliminary data on laser-cut
nipple units indeed support this conten¬
tion (O.P.M., unpublished data, 1989).
A decrease in ventilation during nip¬
ple feeding is well documented in both
term and preterm infants.4,5 Among pre¬
term infants, the reduction in ventila¬
tion is greater in the more premature
groups.5 Even apnea and bradycardia
may occur in some of these infants.6 Im¬
maturity of the central nervous system
presumably is responsible for the above
observation. Alternatively, the high
milk flow present in these studies may
have contributed, at least in part, to the
adverse effects mentioned above. If this
assumption is true, the occurrence of
apnea and bradycardia can be reduced
significantly by decreasing the milk
flow.
Our results raise several questions.
Do we need nipple units specially de-
signed for premature infants? If so, how
should these differ from nipple units de¬
signed for term infants? How much vari¬
ability in flow characteristics is accept¬
able among nipple units designed for a
specific group of infants, such as term or
preterm infants?
The present study documents that
the wide variability in flow characteris¬
tics of the nipple units is due primarily
to differences in the size of feed holes
and suggests the need for design
changes to reduce the observed var¬
iability.
The author thanks Curtis P. Cowen and Mead-
Johnson Nutritional Division for their technical as¬
sistance and Linda Bodiford for secretarial
assistance.
References
1. Mathew OP. Nipple unit for newborn infants:
a functional comparison. Pediatrics. 1988;81:688\x=req-\
691.
2. Duara S, Rufenbergh L, Bancalari E. Vary-
ing feeding container rigidity: effect on intake and
ventilation in preterm infants. Pediatr Res.
1985;19:340.
3. Mathew OP. Regulation of breathing pattern
during feeding: role of suck, swallow and nutrients.
In: Mathew OP, Sant' Ambrogio G, eds. Respira-
tory Function of the Upper Airway. New York,
NY: Marcel Dekker Inc; 1988:535-560.
4. Mathew OP, Clark ML, Pronske ML, Luna-
Solarzno HG, Peterson MD. Breathing pattern and
ventilation during oral feeding in term infants J
Pediatr. 1985;106:810-813.
5. Shivpuri CR, Martin RJ, Carlo WA, Fanaroff
AA. Decreased ventilation in preterm infants dur-
ing oral feedings. J Pediatr. 1983;103:285-289.
6. Mathew OP. Respiratory control during nip-
ple feeding in preterm infants Pediatr Pulmonol.
1988;5:220-224.