Influences on Breastfeeding Outcomes for Healthy Term and Preterm/Sick Infants

Abstract

Objective:

This study investigated breastfeeding outcomes and factors associated with duration of breastfeeding in cohorts of healthy term and sick/preterm infants.

Study Design:

Follow-up was conducted of 87 sick/preterm and 92 term healthy infants during the 9 months following discharge from two Western Australian hospitals.

Results:

When corrected for gestational age, breastfeeding duration did not differ between the entire cohort of sick/preterm and term healthy groups. Multivariable analysis showed that early cessation of breastfeeding was more likely for infants born at <33 weeks of gestation (hazard ratio [HR ]=2.05, 95% confidence interval [CI]=1.14–3.69), mothers who had previously breastfed for <6 months (HR=3.67, 95% CI=2.12–6.37), and for mothers who perceived breastfeeding to be important rather than very important (HR=2.58, 95% CI=1.59–4.20).

Conclusions:

Maternal perceptions of breastfeeding as “important” rather than “very important” and previous breastfeeding duration of <6 months are negatively associated with breastfeeding duration; these factors can be identified and addressed during pregnancy and in the postnatal period. Preterm infants born at <33 weeks of gestation have a shorter breastfeeding duration than those born at a later gestation.

Introduction

The health benefits of breastmilk and breastfeeding for both mother and infant have been reported extensively throughout the literature, with particular benefits noted for preterm and sick infants including improved neurodevelopment and protection from infection.^1,2 Australia has a high breastfeeding initiation rate, with 92% of term infants breastfed at birth and 87% breastfed at hospital discharge.¹ Mothers of preterm and sick newborn infants have reportedly lower rates of breastfeeding on hospital discharge; these rates vary widely among both individual neonatal units and countries with a published range of 49–91%.^2,3 The World Health Organization guidelines recommend exclusive breastfeeding for 6 months, with the introduction of complementary foods thereafter and continuation of breastfeeding for 2 years or beyond.⁴ Despite high initiation rates, there is a decline in breastfeeding rates over time.⁵ Australian statistics fall short of the national goal of 80% breastfeeding at 6 months,⁶ with 56% of infants receiving any breastmilk and 12% exclusively breastfed at 6 months.¹ Breastfeeding duration may be reduced in preterm and sick infants who have been admitted to a neonatal unit.^7,8 Reported breastfeeding rates at 6 months range from 31% to 60% for preterm infants.^9,10

Several factors have been found to influence breastfeeding duration, including sociodemographic, biomedical, and psychosocial factors. Sociodemographic factors such as maternal education and income and biomedical factors such as maternal smoking and infant and maternal health status influence breastfeeding duration in term, healthy infants,^8,10 as do psychosocial factors such as attitudes towards breastfeeding and the partner's infant feeding preferences.^11,12

Limited research has explored factors that impact on breastfeeding outcomes for preterm and sick newborn infants following hospital discharge.¹³ Socioeconomic status has been shown to influence breastfeeding duration for very preterm infants, with lower socioeconomic status being associated with earlier cessation of breastfeeding.¹⁴ There is some evidence that multiparous mothers of preterm and sick newborn infants breastfeed for a shorter duration than their primiparous counterparts.¹⁵ Exclusivity of breastfeeding appears to influence breastfeeding duration, with cessation of breastfeeding occurring earlier in those infants fed both breastmilk and formula at the time of discharge from the neonatal nursery.¹⁶ Preterm and sick newborn infants may face multiple obstacles in the quest to continue breastfeeding beyond the hospital stay, yet this area has not been well studied. In order to assist families to achieve the best possible nutrition, health, and developmental benefits conferred by breastfeeding, it is important to understand the specific challenges integral to this population.

The aim of this study was to investigate breastfeeding duration and factors associated with continued breastfeeding in two Western Australian cohorts of healthy term and preterm and sick newborn infants.

Subjects and Methods

A prospective observational longitudinal study was conducted to describe and compare breastfeeding influences and duration of a cohort of preterm and sick infants with a cohort of healthy term infants across the first 9 months following hospital discharge.

Recruitment was carried out by the first and second authors within 48 hours of the infant's discharge from one pediatric and one women's hospital in Perth, WA, Australia between October 2006 and July 2007. Eighty-seven mothers of preterm and sick infants were recruited from the neonatal nurseries that provide specialist care for newborn infants with complex and critical illness. “Preterm/sick infants” were defined as infants admitted for ≥5 days for the management of morbidities related to preterm birth or other neonatal medical or surgical conditions including meconium aspiration, sepsis, and gastrointestinal and cardiac defects. Of this group of infants, 57% were born preterm, that is <37 weeks of gestation. Ninety-two mothers of healthy term infants were recruited from the postnatal wards of the women's hospital, a Baby Friendly Health Initiative–accredited hospital. Inclusion criteria for both groups were that the mothers were breastfeeding and/or expressing breastmilk for their infant, intended to continue feeding breastmilk following the infant's discharge from hospital, had to be 18 years or older, read and spoke English, and were able to be contacted by telephone. Infants who were to be transferred from the hospital's neonatal unit to another hospital prior to discharge home were excluded as the study could not control for discrepancies in hospital practices that may influence breastfeeding outcomes.

Participants were provided with an information sheet and signed a consent form that was approved by the Ethics Committees of the participating hospitals. Mothers were assured of anonymity and confidentiality and were aware that they could withdraw from the study at any time without repercussions for the care of themselves and their infants.

Basic demographic data were collected, and a breastfeeding questionnaire was administered on recruitment into the study. The breastfeeding questionnaire, developed for the purpose of this study, included questions about the timing of the decision to breastfeed, duration of previous breastfeeding experience, perceived importance of breastfeeding, perceived partner support of breastfeeding, and intended breastfeeding duration. Timing of first contact with the breast, defined as the infant's first contact with the breast regardless of whether the infant actually breastfed at that time, and timing of the first breastfeed were recorded. Mothers also completed the Breastfeeding Self-Efficacy Scale—Short Form (BSES-SF)¹⁷ at this time.

The BSES-SF measures a woman's degree of confidence in her ability to breastfeed and is based on Bandura's social cognitive theory.¹⁷ The BSES-SF has demonstrated adequate psychometric properties including a Cronbach's α coefficient of 0.94. It contains 14 items that are scored using a 5-point Likert scale (1=not at all confident to 5=very confident). Scores are summed to produce a range from 14 to 65, with higher scores representing higher levels of breastfeeding self-efficacy.

Follow-up data were collected by telephone interview at 2 weeks, 6 weeks, 3 months, 6 months, and 9 months following discharge home. At each interview, the mother was asked about her current infant feeding method, with “breastfeeding” recorded if the infant was receiving any breastmilk (including expressed breastmilk) and “not breastfed” recorded if no longer receiving breastmilk. Breastfeeding mothers completed the BSES-SF and were asked about their breastfeeding experience and use of other foods and fluids. For infants receiving only breastmilk in the first 6 months, the use of bottles to feed expressed milk was recorded.

Statistical methods

Descriptive statistics for demographic data were based on frequency distributions and medians, interquartile ranges (IQR), and ranges for categorical and continuous data, respectively. Univariate analysis of demographic data between healthy term and preterm/sick groups included χ² and Fisher exact tests for categorical comparisons and Mann–Whitney tests for continuous outcomes. Duration of breastfeeding was estimated using Kaplan–Meier probability estimates, where those who were breastfeeding at completion of the study or lost to follow-up were censored. Duration of breastfeeding was defined as the number of days until the mother reported breastfeeding had stopped or from date of birth to the final follow-up date if the mother was still breastfeeding at the end of the study (censored cases). For cases lost to follow-up, breastfeeding duration was calculated from the date of birth until the last follow-up date when information was available and was censored if still breastfeeding at that time. Individual characteristics were assessed for their association with duration of breastfeeding using partial likelihood ratio tests. Multivariable analysis was conducted using Cox proportional hazards regression modeling to investigate the simultaneous effect of factors that may influence breastfeeding duration. Covariate effects were summarized using hazard ratios (HRs) and their 95% confidence interval (CI). SPSS statistical software (version 15.0, SPSS, Chicago, IL) was used for data analysis. All hypothesis tests were two-sided, and a p value of<0.05 was considered statistically significant.

Based on a breastfeeding rate of 70% in the early postpartum period,¹⁸ a total sample size of 70 in each of the preterm/sick and healthy term groups achieved 86% power to detect a 25% difference in breastfeeding rates between groups with an adjustment made for other explanatory variables. To account for an attrition rate of approximately 30% in the neonatal period, the total sample size was increased to 180 (Power Analysis and Sample Size [PASS] statistical software, NCSS, Kaysville, UT, 2005).

Results

In total, 193 mothers and 204 infants were recruited for this study. Of the 231 mothers approached, 38 declined participation in the study; the main reason given was that the mothers preferred not to commit to ongoing follow-up while caring for a young infant. Fourteen mothers of multiple birth infants were excluded from the analysis because of inadequate numbers available to analyze the unique challenges faced by mothers of twins.¹⁹ Data from 179 mothers of singleton birth infants were analyzed, of whom 87 were recruited from a population of preterm/sick infants and 92 were recruited from a population of healthy term infants. The majority of mothers were recruited from the tertiary-care women's hospital's postnatal wards and neonatal nurseries (n=134) with the remaining recruited from the tertiary-care pediatric hospital's neonatal nursery. Losses to follow-up at the 9-month time point were 14 of 92 (15%) and 14 of 87 (16%) in the healthy term and preterm/sick groups, respectively.

Within the preterm/sick group, 57% of the infants were born preterm, and two infants (2.3%) were born postterm (>42 weeks of gestation). Univariate Kaplan–Meier survival analysis showed duration of breastfeeding did not differ between preterm infants and sick infants born ≥37 weeks of gestation (log-rank p value=0.178).

The median gestations for the healthy term and preterm/sick groups were 39⁺⁵ weeks (range, 37–42 weeks) and 35 weeks (range, 23⁺⁴–42⁺¹ weeks), and the median lengths of stay were 4 days (range, 2–7 days) and 21 days (range, 7–173 days), respectively. Birth weight was not recorded as it has not been previously associated with breastfeeding duration. There were no differences between groups with regard to maternal age, highest education level, marital status, ethnicity, and socioeconomic status (Table 1). Mothers in the healthy term group were more likely to attend breastfeeding education classes than those in the preterm/sick group (59% vs. 29%, p<0.001). Both groups intended to breastfeed for a median time of 52 weeks, but a higher proportion of mothers in the preterm/sick group intended to breastfeed for longer than 52 weeks (p=0.029). The timing of decision to breastfeed (before pregnancy vs. during/after pregnancy; 79–82% decided to breastfeed before pregnancy), the perceived importance of breastfeeding (80–86% perceived breastfeeding to be very important), and level of partner support (89–90% had very supportive partners) did not differ between groups (p values>0.05). The BSES-SF score at discharge was lower for the healthy term group (median, 57; IQR, 48–63) than for the preterm/sick group (median, 59; IQR, 53–66) (p=0.031).

Table 1.

Demographic and Breastfeeding Background Stratified by Healthy Term and Preterm/Sick Groups

	Healthy term (n=92)		Sick/preterm (n=87)
	Number	%	Number	%	p value
Age <30 years	41	45	40	48	0.69
Highest education level
High school	32	35	33	39	0.84
TAFE/diploma	31	34	26	31
Tertiary	29	32	26	31
Married/de facto	83	90	82	97	0.10
White	76	84	71	86	0.71
SEIFA group 4–6^a	33	39	31	41	0.75
Primiparous	50	54	47	55	0.97
Previously breastfed	42	46	40	47	0.85
Duration (weeks)^a	38	13–66 [1–156]	32	17–52 [1–130]	0.49
Timing of decision to breastfeed
Before pregnancy	75	82	67	79	0.65
During pregnancy	17	19	18	21
Attended education classes	54	59	25	29	<0.001
Importance of breastfeeding^b
Very important	74	80	71	86	0.37
Important	18	20	12	15
Partner support for breastfeeding^b
Very supportive	80	90	75	89	0.90
Supportive	9	10	9	11
Intended duration of breastfeeding (weeks)^a	52	26–52 (8–156)	52	32–78 (6–108)	0.03
LUSCS birth	39	43	30	36	0.34
Gestation at birth (weeks)
<33	—	—	32	39	< 0.001
≥33	92	100	51	61
First breastfeed
<2 hours	62	67	25	30	< 0.001
≥2 hours	30	33	58	70
First contact with breast
<2 hours	62	67	25	30	< 0.001
≥2 hours	30	33	58	70
Expressing at discharge	35	38	58	67	< 0.001
BSES-SF score at discharge^a	57	48–63 [25–70]	59	53–66 [23–70]	0.03

Column percentages may not sum to 100 because of rounding.

Data represent median, interquartile range (Q1, Q3), and range [minimum–maximum].

One case reported breastfeeding not important, and one case reported partner not supportive. In each instance, the case was excluded in the χ² test.

BSES-SF, Breastfeeding Self-Efficacy Scale—Short Form; LUSCS, cesarean section; SEIFA, Index of Relative Socio-Economic Disadvantage¹⁹ grouped 1–6 by postal code, with 1 referring to the least disadvantaged and 6 referring to the most disadvantaged socioeconomic group (thus SEIFA group 4–6 represents the most disadvantaged quartile in this sample); TAFE, technical and further (non-tertiary) education.

At discharge, of those fed only breastmilk, one of 87 (1%) healthy term infant and 25 of 83 (30%) of sick/preterm infants received some expressed milk feeds via a bottle as well as feeding directly at the breast. The proportion of infants fed some expressed milk feeds increased in both groups over the early months following discharge (Table 2). Reasons for bottle use were not recorded.

Table 2.

Proportions Feeding Breastmilk Only and Using Bottles by Healthy Term and Preterm/Sick Groups

	Healthy term infants				Preterm/sick infants
	Breastmilk only		Any bottle use as % breastmilk only		Breastmilk only		Any bottle use as % breastmilk only
Time	Numbers	%	Numbers	%	Numbers	%	Numbers	%
At discharge	87/92	95%	1/87	1%	83/87	95%	25/83	30%
2 weeks	71/87	83%	23/71	32%	57/82	70%	27/57	47%
6 weeks	64/83	77%	23/64	36%	44/83	53%	22/44	50%
3 months	51/82	62%	28/51	56%	33/82	40%	11/33	33%
6 months	43/76	57%	19/43	44%	30/80	38%	7/30	23%

Breastfeeding rates did not differ between groups at 2 weeks after discharge with 97% (84 of 87) and 93% (76 of 82) of mothers breastfeeding in the healthy term and preterm/sick groups, respectively (p=0.318). At 6 months after discharge, breastfeeding had decreased to 70% (53 of 76) and 54% (43 of 80) in the healthy term and preterm/sick groups (p=0.04). At 9 months after discharge, breastfeeding had decreased to 55% (43 of 78) and 37% (27 of 73) in the healthy term and preterm/sick groups (p=0.025). The highest rates of stopping breastfeeding occurred between 2 and 6 weeks after discharge (16%) for the preterm/sick group and between 6 and 9 months after discharge for both the healthy term and preterm/sick groups (15% and 17% reduction, respectively). Univariate analysis showed breastfeeding duration did not differ between the groups with a median duration pf 301 days (Q1=119 days) for the healthy term group compared with 224 days (Q1=91 days) for the preterm/sick group (p=0.092) (Fig. 1). Assessment of maternal characteristics and their univariate association with duration of breastfeeding showed the following characteristics were significantly associated with a shorter duration of breastfeeding: duration of previous breastfeeding ≤6 months (p<0.001), first contact with breast ≥2 hours after birth (p=0.001), use of nipple shield at discharge (p=0.001), nonattendance at breastfeeding education classes (p=0.007), first breastfeed ≥2 hours after birth (p=0.010), preterm birth (p=0.012), intention to breastfeed <6 months (p=0.014), highest level of education achieved being high school, technical college, or other non-tertiary education (p=0.026), and BSES-SF score at discharge (p=0.026). Maternal age <30 years and antidepressant use at discharge approached significance with p values of 0.061 and 0.069, respectively.

FIG. 1.

Comparison of breastfeeding duration in preterm/sick infants born very preterm (<33 weeks of gestation) or preterm (≥33 weeks of gestation) and healthy term infants.

Multivariable analysis showed duration of breastfeeding did not differ between the healthy term and preterm/sick groups (p=0.828) after accounting for gestation <33 weeks at birth, previous breastfeeding experience, and maternal attitude toward importance of breastfeeding (Table 3). Early cessation of breastfeeding was more likely for mothers who had previously breastfed for less than 6 months than for mothers who had never breastfed (HR=3.67, 95% CI=2.12–6.37, p<0.001). All women who had never previously breastfed were primiparous. Duration of breastfeeding did not differ for those who had previously breastfed for 6 months or more and primiparous mothers (p=0.980). Early cessation was also more likely for mothers who perceived breastfeeding to be important rather than very important (HR=2.58, 95% CI=1.59–4.20, p<0.001) and for mothers of infants born at <33 weeks of gestation compared with those born ≥33 weeks of gestation (HR=2.05, 95% CI=1.14–3.69, p=0.017). Intention to breastfeed (p=0.430), BSES-SF score at discharge (p=0.080), and maternal age (p=0.171) were not significantly associated with breastfeeding duration in the final model. There were no significant interactions found between group and maternal or pregnancy characteristics.

Table 3.

Association of Covariates and Breastfeeding Duration

	HR	95% CI	p value
Group
Term healthy	1.00
Preterm sick	1.06	0.64–1.75	0.828
Previous breastfeeding duration
Not breastfed before	1.00
≥6 months	1.01	0.61–1.66	0.980
<6 months	3.67	2.12–6.37	< 0.001
Importance of breastfeeding
Very important	1.00
Important	2.58	1.59–4.20	< 0.001
Gestational age at birth (weeks)
≥33	1.00
<33	2.05	1.14–3.69	0.017

CI, confidence interval; HR, adjusted hazard ratio.

In this study population, all multiparous mothers had initiated breastfeeding with previous children. A significant relationship was found between previous breastfeeding duration and intended duration of breastfeeding for the current lactation. An intention to breastfeed for 6 months or more was reported by 94% of those who had previously breastfed for 6 months or more, 71% of those who had not previously breastfed, and 44% of those who had previously breastfed for less than 6 months (p<0.001).

Discussion

In this study, the overall breastfeeding duration of preterm and sick newborn infants was similar to that of healthy term infants; this finding is supported by a Swedish study that compared breastfeeding outcomes of preterm and sick nursery infants with those of the healthy term population.²⁰ Further analysis revealed that breastfeeding duration for very preterm infants (i.e., those born at <33 completed weeks of gestation) was shorter than for those born at a later gestation. This finding is supported by previous research and clinical observations that very preterm infants experience greater feeding difficulties than their more mature counterparts.¹⁴ Breastfeeding duration data were collected retrospectively, with data collection intervals ranging from 2 to 14 weeks; we acknowledge it is possible this created a source of potential recall bias. Also, it is important to note that for the sick/preterm group, reported rates of breastfeeding at set time points following discharge are not comparable with duration of breastfeeding (as shown in Fig. 1) as these infants may have been receiving breastmilk for several weeks or months prior to discharge.

Several biomedical and psychosocial factors have been shown to impact on feeding ability in very preterm infants and so may contribute to shorter breastfeeding duration. Medically stable preterm infants are usually discharged home soon after demonstrating the ability to suck all feeds, which is typically achieved prior to 40 weeks corrected gestational age. There is evidence that preterm infants can establish full breastfeeding by around 36 weeks of gestation.²¹ However, the findings are not generalizable to very preterm infants as the study infants were free of major medical complications and most were born at >33 weeks of gestation. For the very preterm infant, the ability to produce adequate intraoral vacuum and coordinate sucking, swallowing, and breathing may not be fully developed at the time of discharge,²² resulting in suboptimal breast attachment and inadequate milk transfer. Feeding may be further complicated by biomedical factors such as the effects of major intracranial bleeding (grade III/IV intraventricular hemorrhage, periventricular leukomalacia) and chronic lung disease,²³ common complications in very preterm infants that were recorded within our study sample. These feeding issues may in part account for the high proportion of sick/preterm infants receiving some expressed milk feeds via a bottle at discharge.

It appears that a woman's first breastfeeding experience has a major influence on intended and actual duration in subsequent lactations. In this study, mothers with a history of stopping breastfeeding within 6 months of birth had a 3.67 times greater risk of early cessation of breastfeeding than first-time mothers and those with previous breastfeeding duration greater than 6 months. There was no difference between the latter groups with regard to breastfeeding duration. Our findings confirm evidence that previous breastfeeding duration is predictive of subsequent breastfeeding duration.^24,25 It is possible that early cessation of breastfeeding may result from unresolved breastfeeding challenges rather than intention. A woman's previous inability to achieve her breastfeeding duration goal could undermine her confidence in her ability to breastfeed subsequent children and may be reflected in shorter intended and actual duration in subsequent lactations. This group of women at high risk of early cessation of breastfeeding can be identified early and may benefit from an opportunity to reframe previous breastfeeding experiences alongside targeted interventions.

This is the first article to report maternal perception of the importance of breastfeeding as a significant influence on breastfeeding duration. A Canadian study of breastfeeding outcomes in very low birth weight infants reported that the majority of mothers and their partners rated breastfeeding as very important.²⁶ However, it is unclear whether importance of breastfeeding was considered a covariate when analyzing factors associated with breastfeeding duration. Our study showed that maternal ratings in response to the simple question “How important is breastfeeding to you?” were associated with breastfeeding duration. At the time of hospital discharge, mothers who rated breastfeeding as “important” were 2.5 times more likely to stop breastfeeding early than those who rated breastfeeding as “very important.” It may be possible to influence a mother's perception of the importance of breastfeeding through tailored breastfeeding education and support.^27,28

We found that mothers of sick and preterm newborn infants were significantly less likely to have accessed breastfeeding information during pregnancy through attending classes or reading about breastfeeding. A proportion of these mothers had given birth before 33 weeks of gestation; they are unlikely to have had the opportunity to attend antenatal breastfeeding classes, which are typically accessed in the last trimester of pregnancy. Regardless of this, accessing breastfeeding information during pregnancy was not associated with breastfeeding duration.

Sociodemographic factors such as maternal age and education were found not to be significant influences on breastfeeding duration. Moreover, recent evidence suggests that psychosocial factors may be more influential on breastfeeding duration than sociodemographic factors.^12,29 This is a promising finding as modifiable factors such as maternal perceptions of the importance of breastfeeding may be influenced by psychosocial and educational interventions. Factors such as maternal body mass index and postnatal depression may influence breastfeeding outcomes but were outside the scope of this investigation.

Evidence suggests that breastfeeding confidence, as measured by the BSES-SF in the early postnatal period, is predictive of breastfeeding duration and exclusivity.^17,30,31 The BSES-SF was associated with duration of breastfeeding at univariate analysis only. Published studies of the influence of breastfeeding self-efficacy on breastfeeding outcomes have either focused on primiparous women or have not quantified and included previous breastfeeding duration in a multivariable analysis. The BSES-SF may be clinically useful for primiparous women. However, our results suggest that for women with previous breastfeeding experience, consideration of previous breastfeeding duration and current perception of the importance of breastfeeding could be an accurate and efficient indicator of breastfeeding outcome.

This is the first published report of BSES-SF use in the preterm/sick population. The median BSES-SF score of 59 for the preterm/sick group was significantly higher than the score of 57 for the healthy term group. Increased breastfeeding confidence in the preterm/sick group may have been facilitated through vicarious experience, education, feedback, and support in the nursery setting. However, greater confidence in breastfeeding ability, as measured through the BSES-SF score at hospital discharge, did not impact on breastfeeding duration for this group.

With limited healthcare resources, it is desirable that timely breastfeeding support is targeted to those who need it most. Significant factors that impact on breastfeeding duration can be identified through asking some simple questions of women during pregnancy or in the early postnatal period: “How important is breastfeeding to you?” and “For how long did you breastfeed your older child/ren?” In order to optimize breastfeeding outcomes, the specific challenges of breastfeeding infants born at <33 weeks of gestation warrant further investigation. With higher rates of stopping breastfeeding occurring in the early months following discharge home, community-based breastfeeding support is justified to assist families in achieving their breastfeeding goals.

Footnotes

Acknowledgments

We are grateful to the families who participated in this study and to Mrs. Sandy Andersen who assisted with recruitment. This study was supported by the Women and Infants Research Foundation and the Western Australian Nurses Memorial Charitable Trust.

Disclosure Statement

No competing financial interests exist.

References

Australian Institute of Family Studies. Growing Up in Australia: The Longitudinal Study of Australian Children. Australian Institute of Family Studies: Melbourne, 2008.

Callen

, Pinelli

. A review of the literature examining the benefits and challenges, incidence and duration, and barriers to breastfeeding in preterm infants. Adv Neonatal Care, 2005; 5:72–88.

Trajanovska

, Burns

, Johnston

. A retrospective study of breastfeeding outcomes in an Australian neonatal intensive care unit. J Neonatal Nurs, 2007; 13:150–154.

World Health Organization. Global Strategy on Infant Feeding and Young Child Feeding. World Health Organization: Geneva, 2002.

Callen

, Pinelli

. Incidence and duration of breastfeeding for term infants in Canada, United States, Europe and Australia: A literature review. Birth, 2004; 31:285–292.

National Health and Medical Research Council. Dietary Guidelines for Children and Adolescents in Australia. Commonwealth Department of Health and Ageing: Canberra, 2003.

Dall'Oglio

, Salvatori

, Bonci

et al. Breastfeeding promotion in neonatal intensive care unit: Impact of a new program toward a BFHI for high-risk infants. Acta Paediatr, 2007; 96:1626–1631.

Ladomenou

, Kafatos

, Galanakis

. Risk factors related to intention to breastfeed, early weaning and suboptimal duration of breastfeeding. Acta Paediatr, 2007; 96:1441–1444.

Collins

, Ryan

, Crowther

et al. Effect of bottles, cups, and dummies on breast feeding in preterm infants: A randomised controlled trial. BMJ, 2004; 329:193–198.

10.

Flacking

, Nyqvist

, Ewald

. Effects of socio-economic status on breastfeeding duration in mothers of preterm and term infants. Eur J Public Health, 2007; 17:579–584.

11.

Scott

, Binns

, Graham

et al. Temporal changes in determinants of breastfeeding initiation. Birth, 2006; 33:37–45.

12.

O'Brien

, Buikstra

, Hegney

. The influence of psychological factors on breastfeeding duration. J Adv Nurs, 2008; 63:397–408.

13.

Dowling

, Shapiro

, Burant

et al. Factors influencing feeding decisions of black and white mothers of preterm infants. J Obst Gynecol Neonatal Nurs, 2009; 38:300–309.

14.

Flacking

, Wallin

, Ewald

. Perinatal and socioeconomic determinants of breastfeeding duration in very preterm infants. Acta Paediatr, 2007; 96:1126–1130.

15.

Flacking

, Nyqvist

, Ewald

et al. Long-term duration of breastfeeding in Swedish low birth weight infants. J Hum Lact, 2003; 19:157–165.

16.

Lessen

, Crivelli-Kovach

. Prediction of initiation and duration of breast-feeding for neonates admitted to the neonatal intensive care unit. J Perinat Neonatal Nurs, 2007; 21:256–266.

17.

Dennis

. The Breastfeeding Self-Efficacy Scale: Psychometric assessment of the Short Form. J Obst Gynecol Neonatal Nurs, 2003; 32:734–744.

18.

Cooke

, Sheehan

, Schmied

. A description of the relationship between breastfeeding experiences, breastfeeding satisfaction, and weaning in the first 3 months after birth. J Hum Lact, 2003; 19:145–156.

19.

Damato

, Dowling

, Standing

et al. Explanation for cessation of breastfeeding in mothers of twins. J Hum Lact, 2005; 21:296–304.

20.

Akerstrom

, Asplund

, Norman

. Successful breastfeeding after discharge of preterm and sick newborn infants. Acta Paediatr, 2007; 96:1450–1454.

21.

Nyqvist

, Sjoden

, Ewald

. The development of preterm infants' breastfeeding behavior. Early Hum Dev, 1999; 55:247–264.

22.

Amaizu

, Shulman

, Schanler

et al. Maturation of oral feeding skills in preterm infants. Acta Paediatr, 2008; 97:61–67.

23.

Lau

, Sheehan

, Shulman

et al. Oral feeding in low birth weight infants. J Pediatr, 1997; 130:561–569.

24.

Kronborg

, Vaeth

. The influence of psychosocial factors on the duration of breastfeeding. Scand J Public Health, 2004; 32:210–216.

25.

Otsuka

, Dennis

, Tatsuoka

et al. The relationship between breastfeeding self-efficacy and perceived insufficient milk among Japanese mothers. J Obst Gynecol Neonatal Nurs, 2008; 37:546–555.

26.

Pinelli

, Atkinson

, Saigal

. Randomized trial of breastfeeding support in very low-birth-weight infants. Arch Pediatr Adolesc Med, 2001; 155:548–553.

27.

Chung

, Raman

, Trikalinos

et al. Interventions in primary care to promote breastfeeding: An evidence review for the U.S. Preventive Services Task Force. Ann Intern Med, 2008; 149:565–582.

28.

Kervin

, Kemp

, Pulver

. Types and timing of breastfeeding support and its impact on mothers' behaviours. J Paediatr Child Health, 2010; 46:85–91.

29.

Donath

, Amir

. ALSPAC Study Team. Relationship between prenatal infant feeding intention and initiation and duration of breastfeeding: A cohort study. Acta Paediatr, 2003; 92:352–356.

30.

Awano

, Shimada

. Development and evaluation of a self care program on breastfeeding in Japan: A quasi-experimental study. Int Breastfeed J, 2010; 5:9.

31.

McCarter-Spaulding

, Dennis

. Psychometric testing of the Breastfeeding Self-Efficacy Scale-Short Form in a sample of Black women in the United States. Res Nurs Health, 2010; 33:111–119.