Determinants of energy, protein, lipid, and lactose concentrations in human milk during the first 12 mo of lactation: the DARLING Study13 Laurie
A Nommsen,
ABSTRACT
Factors
ergy-yielding 9, and
Cheryl
associated
nutrients
12 mo
A Lovelady, with
in human
postpartum
search on Lactation, ples were obtained
M Jane concentrations
milk
in the
were
Bo L#{246}nnerdal, and
of en-
examined
DARLING
at 3, 6,
(Davis
Infant Nutrition and by complete expression
over 24 h. Milk energy density was concentration; both were positively
Heinig,
Area
Re-
Growth) Study. Samof alternate breasts
highly related
correlated to maternal
of ideal related
body weight (%IBW) at 6, 9, and 12 mo and negatively to milk volume at 3 mo and to parity at 12 mo. Milk
protein
concentration
was
6 and 9 mo and positively and %IBW at 9 mo. Milk related orrhea protein
related
related lactose
to nursing concentration
16 wk postpartum milk
but
composition
as body
composition,
during
the
first
not
is more diet, few
before.
some
volume
frequency was
at 6 mo positively
and
parity
during
Am
months.
findings
to maternal
at
later
such
lactation
J C/in
Nutr
milk concentrations study was limited
contrast,
data
from
maternal
only become this hypothesis,
poorly
among
well-nourished
and
dietary
factors
breast
milk,
dietary
populations
maternal fat stores milk-sampling pro-
of lactation (Davis Area
women
the effects
on human-milk
prolonged studies
of
composition
breast-feeding. that extend
into
To explore the latter
are necessary. Our data set from the Research on Lactation, Infant Nutrition
than
lipid,
199l;53:
differences
and
energy
thropometric
in the
in milk.
indicators,
These
gitudinal
intake
changes
ofthe
study
measures
will
Subjects
and
data
DARand of 12 to
lactose,
maternal
from
an-
a subgroup
of
variables such as nursing volume. Within-women loncomposition
the interrelationships
be examined
of protein, include
intake
in breast-milk
and
amounts factors
dietary
the participants, and mother-infant frequency, feed duration, and milk Lactation,
nourished
(3, 4). that
evident with longitudinal
interindividual
457-65.
KEY WORDS
of protein, lactose, or to the first 4 mo of lac-
a positive relationship between lipid concentrations when 24-h are used
stages LING
G Dewey
Growth) Study represents the first such effort that we know to collect representative milk samples throughout the first mo oflactation from a relatively large sample of well-nourished women. In this paper we examine factors potentially related
suggest
factors
In
and their
It is possible
to milk
These
sensitive
tation.
cedures
to milk volume at 6 and 9 mo and to continued amenat 9 mo. In a subsample who completed dietary records, intake was positively associated with lipid concentration
after that
negatively
or dietary intake lipid. However, demonstrate and milk
with lipid percent
Kathryn
during
among
in a separate
the
the
12 mo
four
outcome
were
described
publication.
Introduction There
is great
yielding
nutrients
populations amount
(1). of energy
taking
in similar
equate
energy
to identify
many
as maternal
As a result
there
can
and
protein
for growing
ofenergy,
range
in the
by breast-fed the
researchers
have
of ad-
infants,
protein,
lipid,
to milk
attempted
energy,
it is useful and
to relate of these
lactose
varies
both
samples
No significant Am J C/in Nutr
from
et al (2), who
45 well-nourished
relationships l99l;53:457-65.
were
obtained
lactating found
factors One of samples of the
Printed
in USA.
representative in Texas.
maternal ©
1991
fatness
American
overall
Society
Downloaded from https://academic.oup.com/ajcn/article-abstract/53/2/457/4694250 by guest on 02 February 2018
objectives (5).
document
total
Briefly,
data
tinue
infants only
nursing
their
with
I
From
2
Supported
DARLING
intakes
the first
ofNutrition, grant
for Clinical
Nutrition
patterns
86 CRCR
This
who
planned
of 12 mo were
University
to
of breast-
during
40 would
Address reprint requests to K Dewey, University ofCalifornia, Davis, CA 95616. Received November 21, 1989. Accepted for publication June 6, 1990. 3
designed
cohort
mothers that
was
18 mo oflife.
for a minimum
the expectation
by USDA
growth
breast-feeding
breast-feeding
the Department
study
and
during
infants
Study
longitudinal
for the
Ninety-two
recruited
ofthe this
nutrient
formula-fed
12 mo.
diurnally
women
between
The
presents
feedings. Very few researchers have used a sampling that accounts for this variability. One exception to by Butte
design
previously and
nutrients
have been found. is obtaining milk the concentration lipid,
Study
infants
importance
to concentrations
associations investigations intake, because
methods
well-nourished
be a wide
Given
intakes
of energy-
within
milk.
nutrition
contributor
concentrations
consumed
of milk.
protein
consistent in such of 24-h
this was in a study milk
even
in human
in milk, few the difficulties representative major
milk,
the determinants
Although
and within procedure
in the
in human
volumes and
concentrations such
variability
continue
paper the
first
to coninitially through
ofCalifornia
at Davis.
1-1968. Department
of Nutrition,
457
458
NOMMSEN
the first year. Selection criteria for the breast-fed group were mother and infant both be healthy with no chronic illnesses that
the
ofage
mother
plan
or to feed
the first Human
12 mo. Subjects
fomnia,
not
to introduce
120 mL/d
>
solid
ofother
milk
The protocol for this Review Committee
before
or formula
4 mo
throughout
the
92
50, and 46 respectively.
breast-feeding provided Of those
so because
they
data who
were
be inconvenient.
subjects
by the of Cali-
initially
73,
too
chronic
or found
the
to
upper
middle Average
medication,
the study
infants
had
weaned developed
3 subjects
procedures
either
to
been
given
majority
prepregnant
of subjects
incomes,
pregnancy
weight
percent
Mean maternal were primiparous.
ideal
age
was
Milk
sample
were
and
were
white
gain
was
14.8
body 30.4
weight ± 4.6
cup
at each
The remaining Test-weighing
nonHispanic.
(%IBW)
kg and was
36%
101 ± 14%.
of the
mothers
Home several height the
visits infant while
nearest
were
made
and maternal subjects were 0.5
cm
by use
monthly
from
indices (5). At the first visit, maternal not wearing shoes was measured to
was
performed total
of a steel
tape
measure
and
90#{176} angle against a wall. At each monthly visit maternal was measured on a Norelco digital scale (Stamford, CT)
wooden weight accurate
(in g) was added
during
sampled Total
pump,
period,
it carefully
be fed back infant intake
the
same
24-h and
Milk
test
(7).
intake
At 3, 6, 9, and weighing the
12 mo, breast-milk infant before and
by use of an electronic curate to the nearest
balance 1 g and
secutive
older
weights.
from below Subjects their were 24-h
For
the balance were asked
(Sartorius, programmed
infants
a swing
daily schedule for the test-weigh rescheduled ifthe infant showed milk intake was calculated by
on the
last day,
3) multiplying Because there insensible water
and
milk
intake
ment
factor
determined
by the
IWL
was was
typical
time
for each
by measuring
interval
feed of the after that (in hours),
test weighing we corrected
infant. IWL
First, directly
both summer and winter months in a subsample ofinfants. a total of 342 measurements of 37 infants, an average 0.05
g - kg
.
min’
was calculated,
of
which
was relatively
due to breastan adjust-
proportional for analysis.
Lowry
amount
pour
offa
10-
freeze the sample.
to the infant as desired. from the opposite breast period
to assess the
each
sub-
proportion
pumped as described
and the above.
(% was
infant’s
assay
to the volume pumped Protein was determined
(8), with
bovine
serum
at each by use
albumin
(BSA)
result
in slightly
values
cause
of the
elevated
instability
for total
of human-milk
protein. protein
in longitudinal by a modified
to
However,
be-
standards,
they
studies at the Folch extraction procedure was calculated
present (9, 10)
described by by summing
the values for protein, lactose, and lipid multiplying them by 5.65, 3.95, and 9.25 (12). A subset of samples was also analyzed
concentrations and kcal/g, respectively directly for energy
by the bomb-calorimetry The two methods where (13).
by Garza et al (12). as reported else-
Maternal
dietary
A subsample
Park,
yielded
of mothers
during
one
were
provided
NJ)
method described similar results,
intake
balance
volunteered
of the
to record
test-weigh
periods
a small
electronic
with
accurate
to 2 g and
a tape
their
dietary
at 3, 6, or 9 mo. (Ohaus, recorder
Florham that
they
used to document their food intake over a 3
or required
from weight mined
60,
busy
Another
recruited,
AL
product
study was approved of the University
Davis.
Of
(5).
foods
that and
ET
were
analysis package and food-composition tables the sample size for analysis of relationships intake
and
subjects with
an earlier cross-sectional milk composition were
milk
who data
from
composition,
completed 17 subjects
data
dietary
from
records
participating
study in which dietary intake assessed by the same methods
and breastas outlined
(n in
HUMAN-MILK above
(1 5). Characteristics
drawn
from
identical
the
to those
though
half
lactation,
of the
in the
subjects;
were
were
DARLING
exercised was
who
area,
nearly
study.
strenuously
not
different
therefore,
Al-
we included
of
analysis
was
carried
correlation
relationships
out
coefficients
between
dependent
maternal
variables
concentrations
12 mo
all of them
at each
postpartum.
the
were
used
or infant
of milk
to examine
protein,
four
Independent
package
characteristics
lipid,
of the
SPSSx
with
lactose,
time
points:
variables
the
maternal h, and
weight,
nursing
volume
produced in maternal
on
1 1 - 12 mo
resent
total
three
TSF
dummy returned
used
in the
four
dependent
fat,
and
from
data
for all subjects
not
available
degree
the following
fatness: weight,
prepregnancy %IBW, and
able
“%IBW
regressions was most also
because consistently
collection
4 d of test
use
in multiple
elsewhere.
and
on
the per-
carbohydrate,
in ad-
above (except changes return because these among
(r
this
0.78)
>
were
was
subgroup). was
indicative
selected
between
day
and
mean
for
exhibited of maternal
use
and
infant
the
infant
Stepwise
breast-milk
in multiple
intake
with
will
the
selected
the
during
selected
between
was then
and was on
intake was
relationships
regression
associations
volume
volume
breast-milk
multiple
fatness There
breast-milk
Breast-milk
regressions;
simultaneously
12-mo
regressed
it best reflected current maternal related to milk composition.
weighing.
concentrations
or
dietary-intake
intakes
listed menses
which
of visit”
relationship
the
the
8-9, to rep-
maternal weight and %IBW, and maternal TSF at time of visit. The independent van-
at time
a strong
sample
variables,
5-6,
9-,
were
and
of multicolinearity
among
For
carbohydrate
variables TSF and
In addition,
6-,
variables fat,
and
constructed
analysis.
protein,
intake/
2-3,
by the
dition to the independent in maternal weight and were
day.
between
variables
were
of calories
A high
for
nutrient
Results
of Pearson’s
nursing
However,
correlations
frequency because
had these
two
effects
tively correlated with each other, composition tended to be obscured included
in the
volume
and
regression variables were
was on
found
regression.
% pumped. used
milk the
This Thus,
to examine composition.
variables
were
that
van-
although
the
variable
volume
composition.
variables
were
situation forced the
also combined
When
significant
added
to the
itself
existed
with
forward-entry
posi-
was
only
milk
multiple effect
of these
relationships model
maternal mo
actual
(r
parity
and
milk
variable
their
infants
when
was
maternal
P