BREASTFEEDING MEDICINE

Volume 10, Number 6, 2015

ª Mary Ann Liebert, Inc.
DOI: 10.1089/bfm.2015.0014

Daily Breastmilk Volume in Mothers

of Very Low Birth Weight Neonates:
A Repeated-Measures Randomized Trial of Hand
Expression Versus Electric Breast Pump Expression

Mary M. Lussier,1–3 Elizabeth A. Brownell,1–4 Tracey A. Proulx,1–3 Donna M. Bielecki,1–3

Kathleen A. Marinelli,1–4 Sandra L. Bellini,1,3,5 and James I. Hagadorn1–4

Abstract

Background: There are no randomized trials comparing early exclusive hand expression (HE) with early
exclusive electric pump expression (electric expression [EE]) for milk removal in mothers of very low birth
weight (VLBW) infants.
Subjects and Methods: Mothers of VLBW infants were randomized to exclusively HE or EE for the first 7 days
postpartum. Daily volumes of milk were compared between groups for the first 28 days, adjusting for repeated
measures.
Results: The HE (n = 12) and the EE (n = 14) groups did not differ with respect to age, parity, single versus
multiple gestation, or number of expression sessions per day. There were 728 values for daily milk volumes
in the first 28 days, including 105 HE and 623 EE. Mothers using exclusive HE had significantly ( p < 0.05)
less cumulative daily milk production throughout the first 7 days postpartum compared with exclusive EE.
Mean cumulative milk production among mothers using HE in the first postpartum week remained ap-
proximately half that of those using the electric pump throughout the first 28 days, without evident catch up.
In multivariable analysis, each postpartum day was associated with an adjusted increase of 50 mL of human
milk/day during the first 7 days postpartum and an increase of 13 mL/day between postpartum Days 8 and 28.
After adjusting for repeated measures, number of expression sessions per day, and postpartum day, EE was
associated with an advantage in milk production of 119 mL/day during the first 28 postpartum days compared
with HE.
Conclusions: Compared with mothers using EE, mothers using HE had significantly less cumulative daily milk
production during the first 7 days postpartum. This trend continued after the intervention had been discontinued,
and the great majority of expressions in both groups were with EE. Further research to confirm and expand these
findings is warranted.

Introduction small quantities of enteral nutrition (trophic feedings),

E stablishing milk supply in mothers of very low birth
weight (VLBW) (<1,500 g) infants early in the post-
partum period is challenging but crucial. Low nutrient stores,
a high metabolic rate, and an immature gastrointestinal tract
are barriers to survival and growth for such infants.1,2
However, feeding the preterm infant early, beginning with
improves both gastrointestinal structural and functional
development.3 Oropharyngeal administration of colostrum
every 3 hours for 3 days beginning at 48–96 hours of life
significantly decreases clinical sepsis.4 Using human milk
for oropharyngeal mouth care, early trophic feedings, and
enteral nutritive feedings in this population improves clinical
outcomes.3,4

1
Connecticut Children’s Medical Center, Hartford, Connecticut.
2
The Connecticut Human Milk Research Center, Hartford, Connecticut.
3
Divison of Neonatology and 4Department of Pediatrics, University of Connecticut School of Medicine, Farmington, Connecticut.
5
University of Connecticut School of Nursing, Storrs, Connecticut.

312
EXCLUSIVE HAND VERSUS ELECTRIC EXPRESSION
Despite the advantages of an early human milk diet,
mothers of preterm infants often struggle with delayed lac-
togenesis and inadequate milk volumes.5,6 Subsequently,
insufficient production is the most common reason these
mothers discontinue lactation.7 In order for VLBW infants to
receive mother’s own milk during their hospital stay and
beyond, mothers must express their milk very soon following
delivery and frequently thereafter.6–8 Current recommenda-
tions for mothers who are separated from their infants include
pumping within 6 hours of delivery and then every 3 hours
(eight times in 24 hours).7
For mothers who deliver their preterm infant in a hospital
setting in a developed country, the current standard of prac-
tice is to use a hospital grade (multi-user) electric breast
pump.9 Research regarding milk expression in this popula-
tion has focused on comparing types of electric pumps10,11
or methods of electric expression (EE).12–14 A recent Co-
chrane Review observed 15 out of the 25 (60%) studies that
evaluated pumps or products had support from the manu-
facturers.15 This suggests that research regarding milk expres-
sion has been largely industry driven and therefore focused
on electric pumping. However, mothers separated from their
infants often have difficulty removing colostrum from the
breast with an electric pump.16
Hand expression (HE) may be an alternative means of
expressing colostrum while at the same time providing
stimulation to the breast similar to breast massage. HE may
also be an option for mothers who do not have access to
electric pumps or for women who do not want to be depen-
dent on technology to express milk for their infant.
Mothers of preterm infants can achieve high levels of milk
production by combining the use of EE and HE16 or by al-
ternating between HE and EE in the first 24 hours after de-
livery.17 However, we were unable to locate studies comparing
exclusive HE with exclusive EE in mothers of VLBW infants.
The purpose of this study was to compare 28 days of milk
volumes yielded from exclusive HE and exclusive EE during
the first week postdelivery in mothers of VLBW infants. We
hypothesized that there would be no difference in daily mean
volume in the first 28 days of life between mothers who use
exclusive HE and exclusive EE in the first week postpartum.
Subjects and Methods
The study settings were a postpartum inpatient floor lo-
cated in an inner-city regional care center and a Level IV
inner-city neonatal intensive care unit located within the
same building. In the neonatal intensive care unit setting, if
no contraindication to feeding is present, VLBW infants are
fed gut-priming feeds by 48 hours of life at 10 mL/kg/day.
Advancement of feedings begins on Day 3 or 4 (depending on
birth weight) to 20 mL/kg/day and increasing by 15–20 mL/
kg/day every 24 hours as tolerated.
Mothers 18 years of age or older who delivered infants
weighing £1,500 g and at gestational age £32 weeks were
eligible to participate in this study. Mothers were not eligible
to participate if they spoke a language other than English or
Spanish, were too ill to express milk, had prior breast surgery,
had a recent history of substance abuse, or if their infant was
moribund or transferred to another facility. Informed consent
was obtained from all participants. This study was approved by
the Institutional Review Boards of Hartford Hospital and the
313
Connecticut Children’s Medical Center. As this was a pilot/
feasibility study, no sample size calculations were performed.
Participants were randomized to either the HE group or the
EE group. All participants began milk expression using the
assigned method within 6 hours of delivery. An International
Board Certified Lactation Consultant/neonatal intensive care
unit registered nurse instructed each mother on the proper
technique for milk expression relative to her randomized
allocation. International Board Certified Lactation Con-
sultants used an educational video,18 verbal instruction, and
hands-on support to guide subjects through the first milk
expression. Participants were asked to exclusively use the
assigned method of expression for the first 7 days postpartum,
after which they could use either or both methods. Members
of the research team instructed mothers to express milk every
3 hours for 15–30 minutes and to maintain a written log
documenting the time of expression, method of expression,
and the milk volume obtained at each expression through 28
days postpartum. All mothers were seen daily by an Inter-
national Board Certified Lactation Consultant for the first
postpartum week to review milk expression techniques and to
answer questions.
Statistical analysis
Descriptive analyses were performed using SPSS version
17.0 software (IBM, Armonk, NY). Univariate comparisons of
independent data were performed with Student’s t or the
Mann–Whitney U test for continuous variables and the chi-
squared or Fisher’s exact test for categorical variables. Non-
independence occurred in the data due to repeated daily mea-
sures of milk volume for each participant; therefore analyses of
nonindependent data were performed with univariate or mul-
tivariate hierarchical linear regression using HLM version 7.0
(Scientific Software International, Inc., Lincolnwood, IL).
Median milk production increased quickly over the first 7 days,
whereas thereafter the slope of median production increased at
a slower rate (Fig. 1). Accordingly, multivariate regression
analysis used a mixed-model spline regression approach19 in
which the relationship between postpartum day and breastmilk
volume was modeled with continuous but separate lines for
postpartum Days 1–7 and Days 8–28. All statistical tests were
two-tailed with p < 0.05 considered significant.
Results
Forty mothers consented to participate in this study.
Twenty mothers were randomized to the HE group and 20 to
the EE group. There were no significant differences between
the two groups with respect to age, parity, primigravida sta-
tus, or multiple versus singleton gestation. Method of ex-
pression for each session, number of expression sessions per
day, and volume of mother’s milk obtained per session were
reported for each of the first 28 postpartum days by 26 par-
ticipants, yielding 728 values for daily milk volume; these
formed the cohort for analysis. All participants in the ana-
lyzed cohort adhered to the study protocol and exclusively
expressed milk by only their assigned expression method for
the first 7 days. Thereafter, all participants in the EE group
chose to continue using an electric pump for expression. In
the HE group after the first 7 days, two mothers continued to
use HE through 21 days, whereas the remaining mothers did
their sessions were with the electric pump. In total, 105 (14%)
314 LUSSIER ET AL.

FIG. 1. Milk production (in mL/day) for the entire study cohort (n = 26).

of sessions in the first 28 days were by HE, whereas 623 (86%)
used electric pump expression. For each individual mother, the
expression sessions per day ranged from one to 10.
In univariate analyses, participants in the HE (n = 12) and EE
(n = 14) groups did not differ detectably with respect to age,
parity, or number of milk expression sessions per day (Table 1).
However, compared with mothers using EE, mothers using HE
had significantly less daily milk production (Fig. 2) and cu-
mulative milk production (Table 1) throughout the first 7 days
postpartum. Cumulative volumes at 14 and 28 days were not
statistically different between the two groups. However, me-
dian cumulative milk production among mothers using HE in
the first postpartum week remained approximately half that of
those using the electric pump throughout the first 28 days.
In multivariate hierarchical spline regression adjusted for
repeated measures, interval since birth was strongly associated
with milk production, as each postpartum day was associated
with an adjusted increase of 50 mL of human milk/day during

Table 1. Univariate Comparison of Mothers Assigned to Hand Expression Versus Electric Pump
Hand expression (n = 12) Electric pump (n = 14) p
Age (years) 33 (28, 35) 31 (27, 34) 0.52a
Primigravida 6 (50) 7 (50) 1.0b
Multiple gestation 3 (25) 3 (21) 1.0b
Volume (mL/day) 250 (33, 455) 349 (59, 672) 0.69c
Expression sessions per day 5 (4, 7) 6 (4, 7) 0.99c
Cumulative volume (mL) after
Day 1 0.1 (0, 2.6) 3.7 (2, 22) 0.01a
Day 2 0.25 (0, 11) 14 (3, 72) 0.01a
Day 3 15 (0.3, 41) 70 (23, 107) 0.006a
Day 4 119 (6, 167) 265 (135, 328) 0.005a
Day 5 216 (28, 267) 589 (270, 628) 0.001a
Day 6 359 (59, 594) 933 (430, 1,025) 0.002a
Day 7 456 (104, 866) 1,371 (538, 1,411) 0.003a
Day 14 2,387 (1,196, 3,952) 4,657 (1,829, 5,930) 0.25a
Day 28 7,208 (4,240, 14,279) 13,639 (3,020, 17,770) 0.56a
Cumulative volume (mL) between
Days 8–14 1,789 (957, 3,320) 3,326 (487, 4,565) 0.53a
Days 15–28 4,654 (2,610, 10,712) 9,079 (1,191, 11,317) 0.63a
Data are median (interquartile range) values or n (%) as indicated.
a
By Mann–Whitney U test.
b
By chi-squared test.
c
By univariate hierarchical linear regression.
EXCLUSIVE HAND VERSUS ELECTRIC EXPRESSION 315

FIG. 2. Daily milk volumes (in mL/day) for mothers in the electric pump group (n = 14) and the hand expression group
(n = 12). *p < 0.05, **p < 0.01.

the first 7 days postpartum and an increase of 13 mL/day be- exclusive HE had significantly less cumulative daily milk
tween postpartum Days 8 and 28. Daily milk volume was also production throughout the first 7 days postpartum. Moreover,
significantly associated with number of expression sessions the deficit in milk production among HE mothers continued
per day, with each session adding an adjusted 25 mL to daily after postpartum Day 7, when the study intervention had been
volume. After adjusting for these factors, EE was associated discontinued and the great majority of expressions in both
with an increase in milk production of 119 mL/day compared groups were with EE. Although cumulative volumes after
with HE (Table 2). Primigravida status, multiple gestation, and postpartum Day 7 were not statistically different between the
maternal age were not associated with daily milk volume in two groups, median cumulative milk production among
multivariate analysis. mothers using HE in the first postpartum week remained
approximately half that of those using EE throughout the first
Discussion 28 days, a difference of potential clinical importance that
showed little evidence of correction during Days 8–28. In
This randomized controlled trial (RCT) study found that,
multivariable analysis, EE was associated with a significant
compared with mothers using exclusive EE, mothers using
advantage in daily milk production during the first 28 days
postpartum compared with HE, after adjusting for other im-
Table 2. Factors Associated with Breastmilk portant factors including number of expression sessions.
Volume per Day in the First 28 Days Postpartum, Mothers who experience a complicated peripartum course,
Using Multivariate Hierarchical Spline Regression such as mothers of VLBW infants, are more likely to expe-
rience a delayed onset of milk production.20 Early initiation
Coefficient 95% CI p of milk expression and frequent milk removal are important
factors in the development of an adequate milk supply in this
Adjusted mean volume - 154
(mL/day) population.6–8 One pilot study demonstrated that initiating
Hand expression - 119 - 175, -63 < 0.001 milk expression within the first hour after birth is more ef-
(electric pump expression fective than expressing between 2 and 6 hours.21 Although
as reference) (mL/day) future studies may corroborate these findings and inform
Postpartum day, per day 50 39, 61 < 0.001 changes to clinical practice, the present standard of care is to
Postpartum day, per day, - 37 - 49, -25 < 0.001 begin milk expression within 6 hours of delivery, as was done
Days 8–28 in this study. Initiation and frequency of milk expression
Expression sessions per day, 25 14, 37 < 0.001 were similar between both the HE and EE mothers and
per session therefore did not influence our results. Despite the differences
Results were not significantly different for primigravida status, we report, most mothers in both study groups were able to
multiple gestation, or maternal age. provide enough milk for gut-priming and early enteral
CI, confidence interval. feeding advancement.
316
Several investigators have studied the use of HE as a
method of milk removal. Zinaman et al.22 examined the
effects of various pumping methods on milk yield and pro-
lactin and oxytocin release when compared with infant
suckling in term healthy infants. They documented lower
prolactin responses in the HE group compared with the EE
group and, as in our study, observed significantly lower milk
volumes in the HE group; oxytocin levels were not signifi-
cantly different between groups.22 Flaherman et al.23 mea-
sured milk volumes immediately after breastfeeding in
mothers of term infants randomized to either an electric
breast pump or HE. The median volume of expressed milk
was 0.5 mL for mothers who used HE and 1.0 mL for the EE
group, although the difference in volume between the groups
did not reach statistical significance. However, at 2 months,
mothers assigned to the HE group were more likely to be
breastfeeding than mothers who used the EE method.23
Morton et al.16 investigated whether the use of two manual
techniques, HE of colostrum and ‘‘hands on pumping’’
(breast compression and HE combined with electric pump-
ing), could promote establishment and sustainability of a
sufficient milk supply in NICU mothers. Their results dem-
onstrated that mothers of preterm infants can achieve high
levels of milk production by combining the use of EE and
HE.16
Larkin et al.24 demonstrated that mothers of preterm in-
fants can establish a full milk supply (>700 mL/day) by
combining HE after electric pumping during the first 3 days
postpartum and hands-on pumping after lactogenesis II.
Two studies compared milk volumes in mothers of pre-
mature infants using HE and EE. Ohyama et al.17 studied HE
versus electric pumping in 11 mothers of infants in a tertiary
perinatal center in Japan. Mothers alternated HE with EE
during the study. Mothers yielded twice as much milk when
using the HE technique compared with sessions using the
electric pump in the first 48 hours.17 It is unclear whether or
not the differences in volumes between groups would have
continued beyond the 48-hour period. Slusher et al.25 com-
pared EE and HE techniques in mothers of infants admitted to
a special care nursery in Uganda. Any mother of an infant
admitted to the special care nursery who was too ill to
breastfeed but not on ventilator support was eligible to be
included. No weight or gestational age criteria were used for
study recruitment. Slusher et al.25 used nonrandom, sequen-
tial assignment to allocate mothers to one of three groups:
Group 1, double electric breast pump; Group 2, single non-
electric manual pump; and Group 3, HE. Maternal milk
volumes were measured each day over a 7-day period. Group
1 produced the highest daily milk volume (mean, 647 mL),
followed by Group 2 (mean, 520 mL); Group 3 (mean,
434 mL) produced the lowest daily milk volume of the three
expression methods.25 Our study is similar in methodological
approach in that mothers were allocated to either exclusive
HE or exclusive EE; no subjects used a combination of milk
expression techniques during the study period.
Our study is subject to several important limitations. First,
observations lost to follow-up resulted in missing data.
However, we were unable to detect significant differences
between mothers excluded due to lack of complete data and
those included in the analysis with respect to age, parity,
multiple versus singleton gestation, or expression method.
Given the small sample size, we acknowledge the potential
LUSSIER ET AL.
for group differences between those with and without follow-
up with respect to unmeasured factors.
Second, we did not collect data on subjects who were
approached but did not consent to participate in the study.
Mothers consenting to participate in this study may have been
more familiar or comfortable with the nonstandard method of
milk expression (HE), thus potentially creating a selection bias.
Lastly, we did not include a measurement of maternal
comfort or preference in the design of this study. Flaherman
et al.23 found that although differences in milk volumes be-
tween the HE and EE groups were not significant, mothers in
the HE group were more comfortable expressing milk with
others present and also showed a trend toward increased
comfort breastfeeding with others present. Furthermore,
mothers in the HE group were more likely to still be breast-
feeding at 2 months than mothers in the EE group.23 Because
we did not measure this in our study population, we cannot
ascertain whether or not mothers were more comfortable with
one method versus another or if maternal comfort/preference
affected daily volumes. Despite the absence of a direct
measurement of comfort and/or preference, it is notable that
most mothers (77%) chose to use EE as their expression
method after the 7-day exclusive expression period ended.
This study has several important strengths. The first in-
volves study design. This was an RCT; therefore the distri-
bution of both known and unknown potential confounders was
likely to be similar between the HE and EE groups. Second, we
used rigorous statistical methodology in our analysis, and al-
though there were only 26 subjects included in the analysis,
this was a repeated-measures dataset that included 728 out-
comes. Lastly, all subjects adhered to the study protocol of 7
full days of their assigned expression method with no protocol
violations, further strengthening our study results. The will-
ingness of mothers to use HE for the full 7 days is notable as
the electric pump is believed by most mothers and health
professionals to be the superior if not only option for milk
removal in this patient population in developed countries.
Conclusions
These results suggest that exclusive EE is superior to ex-
clusive HE for establishing milk supply in mothers of VLBW
infants. Further research is needed to confirm and expand
upon these findings. Studies may be warranted examining the
role of HE as an adjunct method in combination with EE for
the VLBW infant in the early postpartum period.
Disclosure Statement
No competing financial interests exist.
References
1. Commare C, Tappenden K. Development of the infant in-
testine: Implications for nutrition support. Nutr Clin Pract
2007;22:159–173.
2. Ditzenberger G. Nutritional support for premature infants
in the neonatal intensive care unit. Crit Care Nurse Clin
North Am 2014;26:181–198.
3. Hamilton E, Massey C, Ross J, et al. Early enteral feeding in very
low birth weight infants. Early Hum Dev 2014;90:227–230.
4. Lee J, Kim H, Jung Y, et al. Oropharyngeal colostrum
administration in extremely premature infants: An RCT.
Pediatrics 2015;132:e357–e366.
EXCLUSIVE HAND VERSUS ELECTRIC EXPRESSION
5. Cregan M, De Mello T, Kershaw D, et al. Initiation of
lactation in women after preterm delivery. Acta Obstet
Gynecol Scand 2002;81:870–877.
6. Hill P, Aldag J, Chatterton R, et al. Comparison of
milk output between mothers of preterm and term in-
fants: The first 6 weeks after birth. J Hum Lact 2005;21:
22–30.
7. Furman L, Minich N, Hack M. Correlates of lactation in
mothers of very low birth weight infants. Pediatrics 2002;
109:e57.
8. Hill R, Aldag J, Chatterton R. Initiation and frequency of
pumping and milk production in mothers of non-nursing
preterm infants. J Hum Lact 2001;17:9–13.
9. Meier P, Engstrom J, Patel A, et al. Improving the use of
human milk during and after the NICU stay. Clin Perinatol
2010;37:217–245.
10. Burton P, Kennedy K, Lucas A, et al. Breast pump design
and milk production: a randomized control trial. Presented
at the American Academy of Pediatrics Annual Meeting,
San Francisco, CA, October 2–5, 2010.
11. Francis J. Maternal/infant lactation characteristics (MILC)
study: A comparison of single electric pumping devices
[abstract]. J Hum Lact 2008;24:107.
12. Groh-Wargo S, Toth A, Mahoney K, et al. The utility of a
bilateral breast pumping system for mothers of premature
infants. Neonatal Netw 1995;14:313–316.
13. Hill P, Aldag J, Chatterton R. Effects of pumping style on
milk production in mothers of non-nursing preterm infants.
J Hum Lact 1999;15:209–216.
14. Jones E, Dimmock P, Spencer S. A randomized controlled
trial to compare methods of milk expression after pre-
term delivery. Arch Dis Child Fetal Neonatal Ed 2001;85:
F91–F95.
15. Becker GE, Smith HA, Cooney F. Methods of milk ex-
pression for lactating women. Cochrane Database Syst Rev
2015;2:CD006170.
16. Morton J, Hall J, Wong R, et al. Combining hand tech-
niques with electric pumping increases milk production
in mothers of preterm infants. J Perinatol 2009;29:757–
764.
317
17. Ohyama M, Harumi W, Hayasaka Y. Manual expression
and electric breast pumping in the first 48 h after delivery.
Pediatr Int 2010;52:39–43.
18. Morton J. Hand Expression of Breastmilk. 2008. Available at
http.//newborns.stanford.edu/breastfeeding/HandExpression.
html (accessed July 14, 2015).
19. Marsh L, Cormier D. Spline Regression Models. Sage,
Thousand Oaks, CA, 2002.
20. Sievers E, Haase S, Oldigs H, et al. The impact of peri-
partum factors on the onset and duration of lactation. Biol
Neonate 2003;83:246–252.
21. Parker L, Sullivan S, Krueger C, et al. Effect of early breast
milk expression on milk volume and timing of lactogenesis
stage II among mothers of very low birth weight infants:
A pilot study. J Perinatol 2012;32:205–209.
22. Zinaman M, Hughes V, Queenan J, et al. Acute prolactin
and oxytocin responses and milk yield to infant suckling
and artificial methods of expression in lactating women.
Pediatrics 1992;89:437–440.
23. Flaherman V, Scott C, Gay B, et al. Randomized trial of
electric breast pump use vs. hand expression in mothers of
infants feeding poorly at 12–36 hours. Presented at the
Pediatric Academic Societies 2010 Annual Meeting, Van-
couver, BC, Canada, May 1–4, 2010.
24. Larkin T, Kiehn T, Murphy P, et al. Examining the use and
outcomes of a new hospital-grade breast pump in exclu-
sively pumping NICU mothers. Adv Neonatal Care 2013;
13:75–82.
25. Slusher T, Slusher I, Keating E, et al. Comparison of ma-
ternal milk (breastmilk) expression methods in an African
nursery. Breastfeed Med 2012;7:107–111.
Address correspondence to:
Mary M. Lussier, BSN
The Connecticut Human Milk Research Center
Connecticut Children’s Medical Center
282 Washington Street
Hartford, CT 06106
E-mail: mlussier@connecticutchildrens.org