G Model

WOMBI 877 No. of Pages 7

Women and Birth xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Women and Birth

journal homepage: www.elsevier.com/locate/wombi

Previous breastfeeding experience and its influence on breastfeeding

outcomes in subsequent births: A systematic review
Yi Huanga, Yan-Qiong Ouyangb,*, Sharon R. Reddinga
a
Wuhan University, Wuhan, China
b
Wuhan University, Wuhan, 155 Donghu Rd., Wuchang District, China

ARTICLEINFO
ABSTRACT

Article history:
Received 15 June 2018 Problem: Despite widely recognized benefits of breastfeeding, the worldwide rate of six months
Received in revised form 31 August 2018 exclusive breastfeeding has not up to the WHO recommendation. Multiparas are considered more
Accepted 10 September 2018 likely to initiate breastfeeding and breastfeed much longer. This, however, is not always the case if they
Available online xxx have unsuccessful previous breastfeeding experience. Therefore, whether and how previous
breastfeeding experience affect subsequent breastfeeding outcomes need to be explored sufficiently.
Keywords: Objective: The objective of this review is to summarize the evidence of association between previous
Breast Feeding breastfeeding experience and subsequent breastfeeding outcomes.
Breastfeeding experience Methods: A systematic search of Embase, Web of Science, PubMed, Cochrane Library and CINAHL
Review databases from inception to March, 2018 is conducted for cohort studies regarding “previous
Parity breastfeeding experience” as an influencing factor to subsequent breastfeeding initiation and duration.
A narrative synthesis is used in this review according to PRISMA and study quality is assessed using
the Newcastle-Ottawa Scale.
Findings: Fifteen articles were eligible for this review. Previous breastfeeding experience was
consistently correlated with subsequent breastfeeding initiation and duration. However, previous short
breastfeeding duration and unsatisfactory experience negatively affected subsequent breastfeeding.
Conclusions: Robust evidence for the effect of previous breastfeeding experience on subsequent
breastfeeding is present. Midwives and breastfeeding/lactation consultants need to provide customized
interventions for mothers according to detailed previous breastfeeding experience in order to improve
breastfeeding initiation and breastfeeding duration.
© 2018 Australian College of Midwives. Published by Elsevier Ltd. All rights reserved.

are inconsistent in some studies. Nowadays, researchers

Statement of significance discover that previous breastfeeding experience is a
cogent factor to influence subsequent breastfeeding
outcomes.
Problem or issue
What this paper adds
The substantial benefits of breastfeeding and current
suboptimal statistics on breastfeeding rate. This review summarizes the evidence and reveals that
the multiparas who have no previous breastfeeding
What is already known experience are at increased risk of not breastfeeding.
Besides, short previous breastfeeding duration and
Parity is one of the factors influencing breastfeeding negative PBE have an unfavorable effect on subsequent
outcomes. However, the effects of parity on breastfeeding

1. Introduction
Abbreviations: PBE, previous breastfeeding experience; BF, breastfeeding; EBF,
exclusive breastfeeding; FBF, full breastfeeding; ABF, any breastfeeding; PBF, Human milk is the best natural food for infants. It can reduce
partial breastfeeding.
* Corresponding author.
infectious morbidity and mortality in infants and children and
E-mail addresses: [email protected] (Y. Huang), reduce the risk of children being overweight and obese during
[email protected] (Y.-Q. Ouyang), [email protected] (S.R. Redding). both childhood and adulthood.1,2 Women who breastfeed are also

https://doi.org/10.1016/j.wombi.2018.09.003
1871-5192/© 2018 Australian College of Midwives. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Y. Huang, et al., Previous breastfeeding experience and its influence on breastfeeding outcomes in subsequent
births: A systematic review, Women Birth (2018), https://doi.org/10.1016/j.wombi.2018.09.003
2 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx

protected against the development of breast and ovarian cancer source of nourishment was breast milk and
and diabetes in later life.3
World Health Organization (WHO) suggests that
breastfeeding should begin within one hour after birth and
continue exclusively for 6 months without additional water,
other liquids or solids.4 In addition, WHO4 has set a goal of
achieving an exclusive breastfeeding (EBF) rate of 50.0% for all
infants at 6 months of age by 2025. However, at present, only
about 42.0% of infants around the world are breastfed within
the first hour after birth and data from 2013–2018 indicate that
41.0% of infants between 0 to 6 months of age are exclusively
breastfed.5
Parity, as one of the factors influencing breastfeeding, has
been studied extensively. Some studies find that multiparas are
more likely to initiate breastfeeding 6 and have longer
breastfeeding duration.7 However, other studies present no
effect or even a negative effect of parity on breastfeeding
initiation 8,9 and duration.10 These inconsistencies in parity and
breastfeeding outcomes result in researchers focusing on
multiparas’ previous breastfeeding experiences (PBE).
Experienced multiparas not only tend to repeat their previous
duration of EBF with subsequent children, but are more likely to
repeat it of immediately preceding child.11 On the contrary,
inexperienced multiparas and primiparas are less likely to
initiate breastfeeding 12 and/or less likely to breastfeed for a
longer period on their next child.13 Until now, no systematic
review has been conducted to determine whether PBE is a
consistent factor influencing subsequent breastfeeding initiation
and duration. Therefore, PBE may be an important factor to be
considered and can offer a new insight into future breastfeeding
promotion.
Given the substantial benefits of breastfeeding and current
suboptimal statistics on breastfeeding rate, a systematic review
is conducted to further understand the effects of PBE on
breastfeed- ing outcomes in subsequent births. The purpose of
this review is to evaluate and summarize the evidences which
report the associa- tion between having breastfed previous
child/children and subsequent breastfeeding initiation and
duration.

2. Methods

2.1. Search strategy

The databases Web of Science, the CINAHL, Embase, Cochrane

Library and PubMed were searched for published articles from
inception to March 2018 by two researchers independently. The
search terms used were (“Breast Feeding” OR Breastfeeding OR
“Milk, Human”) AND (experience OR history OR previous OR
prior OR first OR last OR past OR next OR subsequent OR
consecutive). The MeSH term and free term of breastfeeding was
limited to title and abstract. There is no limitation for the free text
words of PBE extracted from extensive literature review. A review
of the references of retrieved articles to identify additional studies
was also conducted. Next, two researchers independently screened
titles and abstracts for potential eligibility and then screened full
texts to determine their inclusion.

2.2. Selection criteria

Studies were eligible if they met the following criteria: (1)

cohort studies; (2) reporting an association between PBE and
subsequent breastfeeding initiation and/or duration. The types of
breastfeeding were as follows14,15: 1) breastfeeding (BF): infant
received breast milk (directfrom the breastorexpressed); 2)
exclusive breastfeeding (EBF): infantreceived only breastmilk
from a motheror wet nurse, or expressed breastmilk, and no
otherliquidsexceptfordrops orsyrups of vitamins, minerals or
medicines; 3) predominant breastfeeding: infant’s predominant
may have also received water and water-based drinks A total of 1852 references were retrieved from five databases
(sweetened water, teas, etc.), fruit juice, drop and syrups of and 5 articles were retrieved from review of the references of
vitamins, minerals, and medicines; 4) full breastfeeding
(FBF): infant received EBF or predominant breastfeeding; 5)
any/partial breastfeeding (ABF/PBF): infant received breast
milk plus formula; (3) including healthy mothers (without
chronic diseases or severe pregnancy complica- tions) and
healthy infants (without any birth defects or severe diseases;
no required admission to a neonatal intensive care unit);
(4) regarding PBE as having breastfed previous
child/children no matter which one was breastfed, which
type of previous breastfeed- ing was used and how long the
previous breastfeeding duration lasted. Studies were
ineligible for inclusion if PBE was regarded as a vicarious
experience which means seeing others successfully
breastfeeding16; and/or mothers had been breastfed
themselves when they were infants.

2.3. Quality appraisal

The quality of included studies was evaluated using the

New- castle-Ottawa Scale (NOS).17 The NOS consisted of
three parts: selection, comparability and outcome. Each
study can obtain a maximum of 4 points within the
selectionpart (if a study was truly or somewhat
representative of the community; if unexposed cohort was
drawn from the same community as exposed cohort; if PBE
was collected by structured interview and if the study was
prospective) and 3 points within the outcome part (if
breastfeeding data were collected using independent blind
assessment or record linkage; if follow-up duration for EBF,
FBF and predominant breastfeeding was no less than 6 months
and for ABF/PBF was no less than 12 months; if follow up
rate of cohorts was no less than 80%). A maximum of 2
points can be given for the comparability part (in this review,
if a study controlled maternal age, education, employment or
delivery mode, itwould receive 1 point; if it
alsocontrolledotherconfounding factors, it would receive
another 1 point). Thus, a study can receive 0 to 9 points and
was classified as high quality (8–9 points), medium quality
(6–7 points), and low quality (<6 points).

2.4. Data extraction

Data were extracted by two researchers independently to

include: author(s), country, publication year, sample size,
follow- up time, measurement, quality limitations and
findings. If there were insufficient data, the original authors
were contacted.

2.5. Data analysis

Although statistical meta-analysis was a preferred

approach for the synthesis of findings, the heterogeneity
among included studies made it unfeasible regarding
various statistical methods, breastfeeding types, follow-up
duration and PBE definitions. Therefore, a narrative
(descriptive) synthesis was used to provide a descriptive
summary of the included studies. Narrative synthesis was an
approach to systematic review and synthesize the findings
from studies which relied primarily on the use of words and
text to summarize and explain the findings of the synthesis 18.
Included studies were divided into breastfeeding initiation
and duration groups to sum up the results and the primary
outcome in this review was the association between PBE and
subsequent breastfeeding outcomes.

3. Results

3.1. Search outcomes

 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx 3

retrieved articles. After exclusion of 581 duplicate references, and 7 studies were “low quality”12,19,22,23,27,30,32. The most common
1276 were selected for reading of the titles and abstracts. Then quality limitations were no representativeness of selected cohort
186 full- text articles were assessed according to selection group and using self-report to assess PBE status and
criteria, 15 articles were included in the review. All phases of the breastfeeding outcomes. With regard to methodology, twelve
selection process, as well as the number of studies included and studies adjusted confounding factors which were presented in
excluded at each stage, can be seen in the flow chart depicted in Table 2.
Fig. 1.
3.2.1. Breastfeeding initiation
3.2. Study characteristics Positive effect of PBE on subsequent breastfeeding initiation
was reported in 3 studies19,21,23 (Table 1). One study23 stated that
This review identified 15 studies12,19–32, which derived from 10 the odds of subsequent breastfeeding initiation increased by
countries (USA, China, Italy, Iran, Mexico, Denmark, Canada, 11.0% with each additional week of breastfeeding the first child.
Argentina, Brazil and Germany). These studies included a total of Moreover, if experienced mothers were more satisfied with PBE,
16,579 mothers which consisted of 8811 without PBE and 7768 they were 102.0% more likely to initiate subsequent breastfeed-
with PBE. ing.23 On the contrary, if PBE contained unsolved baby’s suck or
Table 1 summarized the characteristics, results and quality latch trouble, the odds of subsequent breastfeeding initiation
limitations of 15 studies. Among these studies, experienced decreased to 14.0%.23 The second study19 reported that 72.3% of
mothers were regarded as: 1) having breastfed a previous infant inexperienced mothers initiated breastfeeding, whereas 97.7% of
>1 month19; 2) having breastfed their last child 20–22; 3) having experienced mothers did so (p < 0.001). The third study21
breastfed their first child no longer than 12 months23; 4) revealed that compared with inexperienced mothers, experi-
having breastfed their previous child/children no matter enced mothers only had a 25.0% probability to not initiate
how long and which one.12,24–32 In terms of study quality, 1 breastfeeding.
study was evaluated as “high quality”31, 7 studies were
“median quality”20,21,24–26,28,29
Fig. 1. Flow chart of articles selection.
4 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx

Table 1
Characteristics of Reviewed Studies (N = 15).

Author (year)
country Sample Assessment methods Findings Quality
size score

Breastfeeding initiation (N = 3)
Schafer et al.23 n = 174 Questionnaires mailed during Each additional week of breastfeeding the first child, the likelihood of initiation with a 3 (a, b, c,
USA 2017 12 months second child increased (OR 1.11 95% CI 1.05, 1.18). More satisfied with breastfeeding the d, e, h)
first child, more likely to initiate breastfeeding with a second child (OR 2.02 95% CI 1.26,
3.29). Previous breastfeeding problems decreased the odds of initiating breastfeeding with
a second child (OR 0.14 95% CI 0.03, 0.46).
Hauff et al.19 USA n = 2733 Questionnaires mailed during Higher rate of experienced mothers initiated breastfeeding than inexperienced mothers 3 (a, b, c,
2014 12 months (97.7%; 72.3%; p < 0.0001). d, e, g)
Kohlhuber et n = 731 Questionnaires mailed at 2, 4, 6 Experienced mothers were less likely not to initiate breastfeeding compared with 6 (a, b, e)
al. Germany
21
and 9 months inexperienced d mothers (OR 0.25 95% CI 0.15, 0.42).
2008
ABF/PBF duration (N = 5)
Hauff et al.19 USA n = 2733 Questionnaires mailed during Experienced mothers had longer ABF duration than inexperienced mothers (43 weeks, 3 (a, b, c,
2014 12 months 15 weeks; p < 0.001). d, e, g)
Mortazaviet al.24 n = 358 Telephone interview monthly in There was no difference between experienced and inexperienced mothers in continuing 6 (a, b, e)
Iran 2015 24 months EBF or PBF at 1 month (OR 2.626, 95% CI 0.750, 9.196).
Bai et al.20 China n = 559 Telephone interview at 1, 2, 3, 6, 9 Compared with previous ABF duration >3 months, the risk of weaning subsequent ABF 6 (a, b, e)
2015 and 12 months was higher in no previous ABF experience (HR 3.24 95% CI 2.37, 4.42), and previous
duration of ABF ≤ 3 months (HR 2.56 95% CI 2.05, 3.20).
Brownell et al.12 n = 2413 Questionnaires mailed during Primiparas and inexperienced multiparas were more likely to cease ABF at 4 weeks (OR 5 (a, b, e,
USA 2012 12 months 1.91 95% CI 1.29, 2.81, OR 4.26 95% CI 2.39, 7.59). g)
Lok et al.25 China n = 2487 Telephone interview at 1, 2, 3, 6, 9, Experienced mothers were less likely to cease ABF compared with inexperienced mothers 7 (a, e)
2018 and 12 months no matter whether they were native-born or immigrant (HR 0.58 95% CI 0.44, 0.76; HR
0.56
95% CI 0.42, 0.74).
EBF duration (N = 6)
Bai et al.20 China n = 559 Telephone interview at 1, 2, 3, 6, 9 Compared with longer previous EBF duration (>2 months), the risk of weaning 6 (a, b, e)
2015 and 12 months subsequent EBF was higher in no previous EBF experience (HR 1.82 95% CI 1.46, 2.26), and
shorter previous EBF duration (≤2 months) (HR 1.65 95% CI 1.29, 2.10).
Flores et al.26 n = 262 Home visit at 5 and 30 days and Experienced mothers increased subsequent EBF duration by 0.5 months, compared with 6 (a, b, e)
Mexico 2005 then monthly up to 6 months inexperienced mothers (Coefficient 0.5, 95% CI 0.06, 0.93).
Hauff et al.19 n = 2733 Questionnaires mailed during Experienced mothers had longer EBF duration than inexperienced mothers (15.3 and 3 (a, b, c,
USA 2014 12 months 7.4 weeks, p < 0.001). d, e, g)
Brownell n = 2413 Questionnaires mailed during Compared with experienced multiparas, EBF cessation at 4 weeks was more likely to 5 (a, b, e,
et al.12 USA 12 months occur in primiparas but not in inexperienced multiparas (OR 1.45 95% CI 1.05, 2.00; OR g)
2012 1.95 95% CI 0.85, 4.49).
Vieira et al.27 n = 1309 Home visits at 1 month Inexperienced mothers were more likely to cease EBF in 1 month compared with 5 (a, d, e,
Brazil 2010 experienced mothers (OR 1.24 95% CI 1.15, 1.43). f)
Mortazavi et al. 24
n = 358 Telephone interview monthly in Inexperienced mothers were more likely to predominant breastfeed than EBF at 1 month 6 (a, b, e)
Iran 2015 24 months compared with experienced mothers (OR 1.935. 95% CI 1.023, 3.659)
BF duration (N = 5)
Colombo et al.28 n = 740 Telephone interviews at 15, 40, Compared with no PBE, positive PBE was a protective factor and negative PBE was a risk 6 (a, e, f)
Italy 2018 and 90 days factor to breastfeeding (OR 2.5 95% CI 1.7, 3.7; OR 0.4 95% CI 0.2, 0.6).
Kronborg et n = 466 Questionnaires at 17 weeks Compared to those who had PBE >17 weeks, mothers whose PBE <5 weeks, whose PBE 5 (a, b, e,
al.22
from f)
Denmark 2004 5 to 17 weeks and primiparas were more likely to cease breastfeeding (HR 7.73 95% CI
2.86, 20.82; HR 3.79 95% CI 1.55, 9.21; HR 3.94 95% CI 1.58, 9.79).
Tarrant et al. 29
n = 1317 Telephone interview at 1, 2, 3, 6, 9, Experienced mothers were less likely to cease breastfeeding at 1 month but not at 3 6 (a, b, e)
China 2010 12 months months (OR 0.36 95% CI 0.22, 0.61; OR 0.9 95% CI 0.43, 1.87).
DiGirolamo30 n = 1665 Questionnaires mailed during 60.4% of experienced multiparas, 30.1% of inexperienced multiparas and 37.7% of 3 (a, b, c,
USA 2005 12 months primiparas chose to breastfeed longer than 20 weeks (p < 0.0001). d, e, g)
Berra et al.31 n = 634 Home visit at 30 days, 6, 12, 24, Compared with mothers whose PBE >6 months, primiparas and mothers whose 8 (e)
Argentina 2003 36, 48 and 50 months PBE ≤ 6 months were more likely to cease breastfeeding (RR 1.39 95% CI 1.12, 1.74; RR 1.64
95% CI 1.32, 2.02).
FBF duration (N = 1)
Clifford et al.32 n = 731 Questionnaires mailed at 1 and Primiparas and inexperienced multiparas were more likely to FBF at 1 week and 6 months 5 (a, b, e,
Canada 2006 6 weeks, 3 and 6 months compared with experienced mothers (OR 3.70 95% CI 2.29, 5.98; OR 30.19 95% CI 13.27, g)
68.69; HR 0.60 95% CI 0.47, 0.76; HR 0.31 95% CI 0.23, 0.42).

Note. EBF, exclusive breastfeeding; ABF, any breastfeeding; PBF, partial breastfeeding; FBF, full breastfeeding; OR, odd ratio; HR, hazard ratio; RR, risk ratio; PBE,
previous breastfeeding experience; a, No representativeness of exposed cohort; b, Uses self-report to assess PBE; c, No control of confounding factors; d, No control one of
maternal age, education, employment or delivery mode; e, Uses self-report to assess outcome; f, Inadequate length of follow up; g, Inadequate follow up rate; h,
Retrospective cohort study.

3.2.2. Breastfeeding duration cost of infant formula was a barrier for some mothers to introduce
Positive effect of PBE on subsequent EBF, ABF/PBF and BF formula
duration was found in 12 studies. 12,19,20,22,24–31 Negative effect of
PBE on subsequent FBF duration was discovered in one study. 32
Five studies reported ABF/PBF duration as an outcome (Table
1). Four of them reported that experienced mothers were more
likely to continue ABF/PBF for 4 weeks or 12 months than
inexperienced mothers.12,19,20,25 The remaining one presented no
significant difference between inexperienced and experienced
mothers in continuing EBF or PBF at 1 month24. In Iran, the high
because free formula had not been given to mothers in recent
years24. As for the characteristics of the participants, four
studies included primiparas and multiparas. 12,19,24,25 The rest
one only included multiparas and indicated that short
previous ABF duration and no PBE negatively affected
subsequent ABF duration compared with long previous ABF
duration20.
Six studies reported a positive effect of PBE on
subsequent EBF duration (Table 1). Among these studies,
two studies stated that experienced mothers were more
likely to continue EBF at 1 or 12 months19,27; one study
indicated that experienced mothers had 2 weeks longer EBF
duration than inexperienced mothers26; one
 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx 5

Table 2 4. Discussion
Adjustments of Included Studies (N = 12).

Author (Year) Confounding factors adjusted in studies Positive association was discovered between PBE and subse-
quent breastfeeding initiation/duration. More specifically, short
a b c d e f g h i j k l
p p p p p p p p previous breastfeeding duration and negative PBE were detrimen-
m Lok et al.24 (2018) tal factors to subsequent breastfeeding duration.
p
p p p p p p Three studies reporting breastfeeding initiation were
Colombo et al.27 (2018)
p p p p p p included in this review. Although a positive effect of PBE was
Mortazavi et al.23 (2015)
p p p p p p p found in all of them, different statistical methods and a variation
Bai et al.19 (2015)
p p p p p
Brownell et al.12 (2012) in the definition of PBE restrict generalization of these findings.
p p p p p p p p p p p p
Vieira
Tarrantetetal.al.
26
(2010)
28
(2010) Thirteen studies reported breastfeeding duration as an outcome.
p p p p p
Kohlhuber et al. (2008)20
p p As a result, experienced mothers were more likely to continue
Clifford et al.31 (2006) EBF, ABF/PBF and BF but not FBF than inexperienced mothers
p p p
Flores et al.25 (2005)
p p p p (primiparas and/or multiparas). In addition, negative PBE and
Kronborg et al.21 (2004)
p p p short PBE duration were detrimental factors to breastfeeding
Berra et al. 30
(2003)
duration. Social culture and government policy would have
Note. a, maternal age; b, maternal education level; c, maternal postpartum effect on subsequent breastfeeding duration in some countries.
employment; d, breastfeeding plan; e, family income; f, parity; g, delivery mode;
Therefore, future studies need to take them into consideration
h, grandmothers’ breastfeeding history; i, breastfeeding problems (nipple pain,
mastitis, cracked nipples); j, maternal smoking status; k, using paci fier; l, feeding when exploring the influence of PBE on subsequent
patterns (on demand or on schedule); m, husband feeding preference. breastfeeding outcomes.
The reasons why PBE can influence subsequent breastfeeding
outcomes had not been reported by included studies in this
study revealed that experienced mothers were more likely to review. After searching related literature, psychological factors
continue EBF rather than predominant breastfeeding at 1 month in relation to breastfeeding were most mentioned. Researches
compared with inexperienced mothers24. In Iran, the inexperi- showed that positive PBE can actively influence subsequent
enced mothers were easily influenced by the opinion of their breastfeeding outcomes directly and indirectly by increasing
older relatives who believed that infants should be fed by sugar experienced mothers’ subsequent breastfeeding attitude,
water and herbal tea to prevent or cure infant colic and confidence, self- efficacy, motivation and intention34–36, but all of
jaundice33. Apart from the four studies above, the fifth found which would be undermined by a negative PBE16,37. The common
that compared with experienced multiparas, primiparas were negative PBE were unsolved breastfeeding difficulties such as
more likely to cease EBF at 4 weeks but there was not significant suck or latch problems, painful nipples, and unfulfilled
difference in inexperi- enced multiparas12. The rest one only expectations of breast- feeding duration in the past22. It was
included multiparas and discovered that both no PBE and short worth noting that even though mothers had negative PBE, if
previous EBF duration had a negative effect on subsequent EBF they knew the long-term benefits of breastfeeding to infants
duration compared with long previous EBF duration20. such as preventing obesity, they were more likely to initiate
Five studies reported a positive effect of PBE on subsequent EBF with their next child than those who had negative PBE but
BF duration (Table 1). However, there also were somedifferences did not know the long-term benefits23. Therefore, it is important
in each study. The first study presented that experienced mothers for midwives and breastfeeding/ lactation consultants to
were less likely to cease BF at 1 month but not at 3 months review a multipara’s difficulties in PBE and offer targeted
compared with inexperienced mothers29. This study was carried guidance and problem-solving strategies to increase their
out in Hong Kong where the maternity leave was offered for a breastfeeding self-efficacy and intention to subsequent
maximum of 10 weeks29, but there were 73.7% of mothers would breastfeeding38. In addition, specific cognitive strategies such as
return to work postpartum in this study. The second study purposeful recalling of positive PBE may increase experienced
indicated that 60.4% of experienced multiparas, 30.1% of multiparas’ self-efficacy39. Moreover, more long-term benefits of
inexperienced multiparas and 37.7% of primiparas chose to breastfeeding also should be emphasized to all mothers. It is
breastfeed longer than 20 weeks (p < 0.001)30. The third study also important to help first-time mothers to prevent and address
reported that compared with no PBE, negative PBE was a early breastfeeding problems38, which not only influence the
detrimental factor and positive PBE was a beneficial factor to BF breast- feeding experience and duration with their first child,
duration28. The fourth study indicated that compared to mothers but also
whose PBE was more than 17 weeks, mothers whose PBE was impact breastfeeding outcomes of future children.
less than 5 weeks had an almost 8 times higher subsequent BF Breastfeeding intention of primiparas and experienced multi-
cessation rate; similarly, the BFcessationratesformotherswhose paras can be influenced by different factors. Bartle and Harvey16
PBE was from 5 to 17 weeks and primiparas were also indicated that formula feeding social norms could negatively
increased22. The last study found that compared with mothers influence primiparas’ breastfeeding attitude. However, breastfeed-
whose previous BF duration was more than 6 months, ing social norms could positively influence multiparas’
primiparas and mothers whose previous BF duration was less breastfeed- ing attitude and negatively influence formula
than 6 months were more likely to cease subsequent BF21. feeding attitude16. In reference to the social influence of others on
One study reported a negative effect of PBE on subsequent a mother’s breastfeeding intention, the role of a grandmother
FBF duration (Table 1). This study found that primiparas and who prefers to breastfeeding had significant positive influence
inexperienced multiparas were more likely to continue FBF at on primiparas’ breastfeeding attitude, subjective norms, self-
1 week and less likely to cease FBF at 6 months compared with efficacy, respectively and negative impact on formula feeding
experienced multiparas 32. This study was conducted in Canada attitude, whereas it only had positive effect on multiparas’
where the provincial government earmarked funds to provide breastfeeding subjective norms16.
all new mothers with a phone call and a home visit from a Therefore, governmental programs should make greater
public health nurse. In this study, primiparas, inexperienced efforts to reduce advertising about infant formula and to create
multiparas and experienced multiparas accounted for 55%, 26% and foster a breastfeeding social environment, which can
and 19% of participants, respectively. Also, participants were include the promotion of breastfeeding and offering public
more likely to be married, older, better educated and wealthier. facilities for women to breastfeed40. In addition, community or
population based breastfeeding intervention including
husbands,
6 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx

grandmothers, other important family members, and health

carefully, especially previous breastfeeding difficulties. It would
professionals in infant feeding discussions may be more
be better to individualize interventions for multiparas of
effective than an individual approach41.
different PBE status and offer them various forms of
Besides, Bartle and Harvey 16 reported that the effect of self-
encouragement. Standard statistics should be obtained
efficacy on breastfeeding continuation was much stronger in
regarding the average duration of PBE in the community, in
experienced mothers than primiparas. Moreover, primiparas
order to standardize long and short duration of PBE for research
were more likely to be influenced by social environment and
purposes. Regarding PBE satisfaction, standard scales can be
culture rather than self-efficacy16. Hence, this necessitated those
used to assess and measure variations in satisfaction/
promot- ing breastfeeding to consider that an intervention based
dissatisfaction rather than relying on subjective reports. Up to
on increasing self-efficacy might produce little effect in
now, almost all of scales are used to assess current breastfeeding
primiparas but bring notable results in experienced multiparas.
experience. Therefore, researchers can consider combination of
On the other hand, understanding the influence of family
qualitative and quantitative methods to explore how to assess
members on a woman’s decision to breastfeed was an important
PBE effectively. There are also some new sights on PBE needing
consideration in supporting primiparas.
to be explored, such as experienced multiparas’ reasons why
Bakoula et al.42 stated that although primiparas had more
ceasing previous breastfeeding and why some experienced
positive intention about breastfeeding than inexperienced multi-
multiparas choose to decrease their current breastfeeding
paras, inexperienced multiparas were more likely to persevere
duration compared to PBE.
in EBF than primiparas because they would feel guilt and grief
after giving formula to their previous child 43. Thus healthcare
5. Conclusions
pro- fessionals can make efforts to turn the positive
breastfeeding intention of primiparas into the action of EBF and
Previous breastfeeding experience is positively associated with
give more information and consultation about EBF to
subsequent breastfeeding initiation, EBF, ABF/PBF and BF
inexperienced multi- paras to help them initiate EBF and
duration. Besides, short previous breastfeeding duration and
breastfeed exclusively for longer periods.
negative PBE have an unfavorable effect on subsequent
breastfeeding duration. As for the possible reasons, positive PBE
4.1. Limitations and strengths
can increase mothers’ attitude, confidence, self-efficacy, motivation
and intention to breastfeed subsequent child. If mothers experience
Although observational studies were the most appropriate
unsolved breastfeeding difficulties in the past, they will have less
design to assess the questions asked in this review, there were
breastfeeding self-efficacy and intention. As for the mothers
some inevitable biases. First, selecting bias and reporting bias
without PBE, social environment and culture may have more effect
existed in included studies because almost all sampling
on subsequent breastfeeding intention than self-efficacy does.
methods were convenient sampling or purposive sampling,
Hence, midwives and breastfeeding/lactation consultants should
which reduced the representativeness of the sample and limited
provide individual prenatal and postnatal breastfeeding support to
the generaliza- tion of the findings of this review. In addition,
mothers.
nearly all PBE status and breastfeeding outcomes were self-
reported by participants. Mothers were prone to overestimate
Competing interests
their previous and current breastfeeding duration, which may
cause reporting bias. Second, seven studies were regarded as
The authors declare that they have no competing interests.
“low quality”. Additional quality limitations of these “low
quality” studies were inadequate length of follow up,
inadequate follow up rate and no control of confounding Funding
factors, which would make the results overestimated. Therefore,
it was recommended that further studies use more advanced The authors received no financial support for the research,
statistical methods, adequate follow up duration and adequate authorship, and/or publication of this article.
follow up rate to explore the association. Third, this review
searched only for published articles, which may lead to Ethical statement
publication bias. Fourth, this review included quantitative
studies but qualitative method was also an important approach An Ethical Statement is not applicable in this article.
to explore PBE in depth and to fill the gap, which resulted in
selection bias. Although limitations existed, some strengths References
should be acknowledged. Firstly, nearly all included studies
1. Horta B.L., Victora CG. Long-term effects of breastfeeding: a systematic review.
were prospective cohort studies, which can measure actual Geneva, Switzerland: World Health Organization; 2013. p. 13–5 ISBN 978 92
subsequent breastfeed- ing outcomes rather than future 4 150530 7.
breastfeeding intention. Secondly, most of the studies had 2. Thomas DG. Is breast best in the west of England? Use of linked routinely
collected data sources to quantify the protective effect of breastfeeding
adjusted common confounding factors, which made the results against infectious gastrointestinal and respiratory illness during the first year
more reliable. Finally, at present, this was the only one to of life. Lancet 2018;384:S78.
systematically explore the association between PBE and 3. Victor CG, Bahl R, Barros AJ, França GV, Horton S, Krasevec J, et al.
Breastfeeding in the 21st century: epidemiology, mechanisms, and lifelong effect.
subsequent breastfeeding outcomes, which provided a new
Lancet 2016;387:475–90.
insight for future studies and interventions. 4. World Health Organization. Infant and young child feeding. 2018. http://www.
who.int/mediacentre/factsheets/fs342/en/ [Accessed 25 September 2018].
5. United Nations Children’s Fund, Division of Data Research and Policy.
4.2. Implications
Global UNICEF Global Databases: Infant and Young Child Feeding, 2018.
6. Kitano N, Nomura K, Kido M, Murakami K, Ohkubo T, Ueno M, et al.
The relationship between PBE and subsequent breastfeeding is Combined effects of maternal age and parity on successful initiation of exclusive
breastfeeding. Prev Med Rep 2016;3:121–6.
significant in this review. These results suggest that midwives and
7. Patil CL, Turab A, Ambikapathi R, Nesamvuni C, Chandyo RK, Bose A, et
breastfeeding/lactation consultants to review a history of a al. Early interruption of exclusive breastfeeding: results from the eight-
mother’s PBE early in pregnancy before offering any interventions. country MAL- ED study. J Health Popul Nutr 2015;34(1):1–10.
Beyond that, a detailed description of PBE duration and the 8. Ekubay M, Berhe A, Yisma E. Initiation of breastfeeding within one hour of
birth among mothers with infants younger than or equal to 6 months of age
mother’s satisfaction with the experience also need to be asked
 Y. Huang et al. / Women and Birth xxx (2018) xxx–xxx 7

attending public health institutions in Addis Ababa, Ethiopia. Int Breastfeed J 26. Flores M, Pasquel MR, Maulén I, Rivera J. Exclusive breastfeeding in 3 rural
2018;13:4. localities in Mexico. J Hum Lact 2005;21(3):276–83.
9. Villar M, Santa-Marina L, Murcia M, Amiano P, Gimeno S, Ballester F, et al. 27. Vieira GO, Martins CDC, Vieira TDO, De Oliveira NF, Silva LR. Factors
Social factors associated with non-initiation and cessation of predominant predicting early discontinuation of exclusive breastfeeding in the first month of life. J Pediatr
breastfeeding in a mother-child cohort in Spain. Matern Child Health J 2018;47 2010;86(5):441–4.
(Suppl. 2):1–10. 28. Colombo L, Crippa BL, Consonni D, Bettinelli ME, Agosti V, Mangino G, et
10. Islam MJ, Baird K, Mazerolle P, Broidy L. Exploring the influence of al. Breastfeeding determinants in healthy term newborns. Nutrients 2018;10 (1):1–
psychosocial factors on exclusive breastfeeding in Bangladesh. Arch Womens 10.
Ment Health 2017;20(1):173–88. 29. Tarrant M, Fong DY, Wu KM, Lee IL, Wong EM, Sham A, et al. Breastfeeding
11. Phillips G, Brett K, Mendola P. Previous breastfeeding practices and and weaning practices among Hong Kong mothers: a prospective study. BMC
duration of exclusive breastfeeding in the United States. Matern Child Health J Pregnancy Childbirth 2010;10(1):27.
2011;15 (8):1210–6. 30. DiGirolamo A, Thompson N, Martorell R, Fein S, Grummer-Strawn L.
12. Tavoulari EF, Benetou V, Vlastarakos PV, Andriopoulou E, Kreatsas G, Linos A. Intention or experience? Predictors of continued breastfeeding. Health Educ
Factors affecting breast-feeding initiation in Greece: What is important? Behav 2005;32(2):208–26.
Midwifery 2015;31(2):323–31. 31. Berra S, Sabulsky J, Rajmil L, Passamonte R, Pronsato J, Butinof M.
13. Brownell E, Howard CR, Lawrence RA, Dozier AM. Delayed onset Correlates of breastfeeding duration in an urban cohort from Argentina. Acta
lactogenesis II predicts the cessation of any or exclusive breastfeeding. J Pediatr Paediatr 2003;92(8):952–7.
2012;161 (4):608–14. 32. Clifford TJ, Campbell KM, Speechley KN, Gorodzinsky F. Factors
14. Labbok M, Krasovec K. Toward consistency in breastfeeding definitions. Stud Fam influencing full breastfeeding in a Southwestern Ontario community:
Plann 1999;21(4):226–30. assessments at 1 week and at 6 months postpartum. J Hum Lact 2006;22(3):292–
15. World Health Organization. Indicators for assessing breastfeeding practices. 304.
Geneva, Switzerland: World Health Organization; 1991. 33. Jalali AK, Sanagoo A, Joibari L. Causes of unsuccessful exclusive
16. Bartle NC, Harvey K. Explaining infant feeding: the role of previous breastfeeding in mother registered in Gorgan health centers. Journal of Gorgan
personal and vicarious experience on attitudes, subjective norms, self- University of Medical Sciences 2001;8:81–6.
efficacy, and breastfeeding outcomes. Br J Health Psychol. 2017;22(4):763. 34. Moimaz SAS, Rocha NB, Garbin CAS, Rovida TA, Saliba NA. Factors
17. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, Tugwell P. affecting intention to breastfeed of a group of Brazilian childbearing women. Women
The Newcastle-Ottawa Scale (NOS) for assessing the quality of non- Birth 2017;30(2):e119–24.
randomized studies in meta-analysis. Appl. Eng. Agric. 2012;18(6):727–34. 35. Dennis CL. Identifying predictors of breastfeeding self-efficacy in the
18. Petticrew M, Roberts H. Systematic reviews in the social sciences. MA: Blackwell immediate postpartum period. Res Nurs Health 2006;29(4):256–68.
Malden; 2006. 36. Tenfelde SM, Finnegan L, Miller AM, Hill PD. Risk of breastfeeding
19. Hauff LE, Leonard SA, Rasmussen KM. Associations of maternal obesity and cessation among low-income women, infants, and children: a discrete time
psychosocial factors with breastfeeding intention, initiation, and duration. Am J Clin Nutr survival analysis. Nurs Res 2012;61(2):86–95.
2014;99(3):524–34. 37. Koskinen KS, Aho AL, Hannula L, Kaunonen M. Maternity hospital practices
20. Bai DL, Fong DY, Tarrant M. Previous breastfeeding experience and and breast feeding self-efficacy in Finnish primiparous and multiparous
duration of any and exclusive breastfeeding among multiparous mothers. Birth women during the immediate postpartum period. Midwifery 2014;30(4):464– 70.
2015;42 (1):70–7. 38. Ouyang YQ, Su M, Redding SR. A survey on difficulties and desires of
21. Kohlhuber M, Rebhan B, Schwegler U, Koletzko B, Fromme H. Breastfeeding breast- feeding women in Wuhan, China. Midwifery 2016;37:19–24.
rates and duration in Germany: a Bavarian cohort study. Br J Nutr 2008;99 39. Dennis CL. Theoretical underpinnings of breastfeeding confidence: a self-
(5):1127–32. efficacy framework. J Hum Lact 1999;15:195–201.
22. Kronborg H, Vaeth M. The influence of psychosocial factors on the duration of 40. World Health Organization. Global strategy for infant and young child feeding.
breastfeeding. Scand J Public Health 2004;32(3):210–6. [118_TD$DIFF][119_TD$DIFF]Geneva, Switzerland: World Health Organiza-
23. Schafer EJ, Campo S, Colaizy TT, Mulder PJ, Ashida S. Influence of tion; 2003. p. 21–2 ISBN 92 4 156221 8.
experiences and perceptions related to breastfeeding one’s first child on 41. Ke J, Ouyang YQ, Redding SR. Family-centered breastfeeding education to
breastfeeding initiation of second child. Matern Child Health J promote primiparas’ exclusive breastfeeding in China. J Hum Lact 2018;34
2017;21(6):1288–96. (2):365–78.
24. Mortazavi F, Mousavi SA, Chaman R, Wambach KA, Mortazavi SS, 42. Bakoula C, Veltsista A, Prezerakou A, Moustaki M, Fretzayas A, Nicolaidou P.
Khosravi A. Breastfeeding practices during the first month postpartum and Working mothers breastfeed babies more than housewives. Acta Paediatr
associated factors: impact on breastfeeding survival. Iran Red Crescent Med J 2015;17 2007;96(4):510–5.
(4):1–8. 43. Gallagher L. Sub-optimal breastfeeding experience and its effects on women,
25. Lok KYW, Bai DL, Chan NPT, Wong JYH, Tarrant M. The impact of factors affecting initiation and duration of breastfeeding in Ireland. University of
immigration on the breastfeeding practices of Mainland Chinese immigrants in Dublin Trinity College; 2012 (Doctoral dissertation).
Hong Kong. Birth 2018;45(1):94–102.