Women’s perspectives on antenatal breast expression: a cross-sectional survey

The United Kingdom (UK) Infant feeding Survey has demonstrated that the prevalence of breastfeeding is higher amongst mothers from managerial and professional occupations, those who left education aged over 18, mothers aged 30 and over, mothers from ethnic minority groups and mothers living in the least deprived areas [5].

A lower prevalence of breastfeeding and poorer breastfeeding outcomes among women who are overweight or obese is well supported by epidemiological studies [6‐13]. Maternal obesity is associated with up to 13% lower breastfeeding initiation rates, and shorter duration of any breastfeeding and exclusive breastfeeding when compared to women of normal weight [13, 14]. This is of particular concern given the rising rate of overweight and obesity across the globe [15], with rates in women aged 20 years or over within the UK in 2013 standing at 57.2% with a body mass index (BMI) ≥ 25 kg/m² and 25.4% with a BMI ≥30 kg/m² [16].

Although psychosocial factors are significantly associated with the lower incidence of breastfeeding among women with a BMI over 25 kg/m², anatomical and physiological alterations may also play a role. Among these is that women who are overweight or obese may have mammary hypoplasia or insufficient glandular tissue as obesity in childhood negatively affects the development of breast glandular tissue [13]. Many of the characteristics experienced by women who are overweight or obese are consistent with this, including reporting stopping breastfeeding due to perceived insufficient supply [17] and being more likely to try to express in the first 2 months postpartum but less likely to have successfully expressed than women with a normal BMI [18]. A high pre-pregnancy BMI is also a predictor of delayed onset lactogenesis II [19], which is the production of copious milk triggered by progesterone withdrawal after the removal of the placenta [13, 20]. This most critical stage of the lactation cycle [21] is more likely to be delayed (occurring more than 72 h after birth) amongst women with a high BMI than women of normal weight [22]. Various underlying physiological reasons for the increased likelihood of delayed lactogenesis II amongst women with obesity are proposed; the impact of increased oedema [13], increased likelihood of medical problems including gestational diabetes, prolonged labour, caesarean section or preterm birth [6, 23], the role of leptin released from adipose tissue which inhibits oxytocin and milk ejection [13], reduced prolactin response to infant suckling [24], and the impact of insulin imbalance [13].

The practice of antenatal breast expression (ABE) has been proposed as a strategy to prevent delayed lactogenesis II in both the academic and consumer literature [25, 26], and as a strategy to overcome the effects of delayed lactogenesis II by ensuring women have a store of expressed milk to prevent the use of formula milk, particularly if the mother has pre-existing or gestational diabetes [27‐29]. ABE is widely recommended in UK maternity units [30‐43]. A recent survey demonstrated that out of the 56 responding maternity units across 9 geographical regions in the UK, 73% offered ABE to diabetic women, 25% offered ABE to women who had risk factors for neonatal hypoglycaemia, and 19% offered ABE to all women [44]. Furthermore, the practice of ABE is promoted by some lactation support websites [29, 45].

ABE involves expressing colostrum from the breast in the antenatal period, however there is little consensus on timing of onset, frequency, duration and method of expression. Recommendations vary within the academic and consumer literature and local UK guidelines on when to commence ABE. A majority of UK maternity units recommend commencing ABE from 36 to 37 weeks (98%) with only 2% recommending commencing ABE at 35,38 or 39 weeks gestation [44], however consumer literature recommends commencing ABE from 32 to 34 weeks of pregnancy [29]. Similarly recommendations regarding how to express ABE vary, including frequency ranging from once a day [35], building up to 4 times a day [32, 34, 39] a minimum of 4 times a day [43] or as often as the woman wants [36, 40] and duration varying from 5 min [35] up to 20 min at a time [32, 38, 39].

Evidence of the effectiveness or underlying mechanism of action of ABE are still not clear, although several studies including a recent large randomised controlled trial have focused on evaluating the safety and efficacy of ABE among women with diabetes [27, 46‐48]. Initial concerns raised in a retrospective cohort study about the safety of ABE in regards to influencing the timing of onset of labour if ABE was commenced prior to 37 weeks’ gestation [49] have not been supported by a large randomised controlled trial of women at low risk of complications with diabetes in pregnancy [48]. This trial found no difference in gestational at birth between those randomised to ABE from 36 weeks’ gestation and women randomised to standard care [48].

Given that ABE is a widely recommended practice in maternity units and on lactation support websites and that limited evaluation of the acceptability of ABE to prepare for breastfeeding has been undertaken, a wide scoping of acceptability and women’s views on this practice merits a focused exploration.

This study was therefore designed to assess the general knowledge of ABE among mothers, their practices surrounding ABE and the acceptability of ABE to them should it be found to be an effective preparation for breastfeeding. It was also aimed to explore any differences in knowledge and acceptability within the overweight and obese subgroup of mothers.

BMI values were calculated from the reported height and weight measurements and grouped into 3 categories; BMI less than 25 kg/m², overweight (BMI of 25–29.9 kg/m²) and obese (BMI of 30 kg/m² or more) [50]. Occupations of the participants were coded using the 3 category National Statistics Socio-economic Classification (NS-SEC) system [51].

Simple thematic analysis was used for the open ended questions by coding the data after familiarisation, and deriving categories and themes inductively.

There were 797 completed surveys. Nineteen responses were removed; eighteen repeat responses and one with implausible demographic responses. A total of 778 responses were coded and taken forward for analysis. The participants completing the survey ranged in age from 19 to 62 (mean 33.3, SD 5.7). Most of the respondents (94.8%) were living in the UK, 2.9% were in North America and a small proportion were living in other European countries or the continents of Asia, Australia and Oceania or South America (2.1%). Further analysis was limited to the 688 participants who were living in the UK, of childbearing age (16–44) and who had given birth to at least one child. Characteristics of the participants are presented in Table 1.

Most of the participants (95.5%) identified their ethnicity as White, 2.3% as Mixed ethnicity, 1.0% as Black, 1.2% as Asian. The wide geographical distribution of UK respondents can be seen in Additional file 1. The BMI of the respondents ranged from 16.7 kg/m² to 66.6 kg/m² (mean 26.8 kg/m², SD 6.0), 43.2% had a BMI of less than 25 kg/m², 32.1% were categorised as overweight and 24.7% were in the obese category.

The largest occupational group category amongst the participants was higher managerial, administrative and professional occupations (52.3%), although 21.9% of participants’ occupations fell into the NS-SEC unclassified category. A total of 46.8% of the participants had given birth to one child, 37.6% had given birth to two children and 15.6% three or more children.

A total of 677 participants had breastfed (98.4%), with 337 of these women (50.0%) mentioning that they were currently breastfeeding at the time of completing the survey. Only 650 of the participants responded to the question about longest length of time breastfeeding a child, which ranged from 0.05 months to 72 months (mean 17.3, median 15.0 (IQR = 7.7–24.0)); 95.2% reported that they were still breastfeeding at 8 weeks, 84.0% were breastfeeding at 6 months, 64.8% were breastfeeding at 12 months and 58.8% breastfed beyond 12 months.

A total of 442 (64.5%) of the respondents had heard of ABE, 56 (8.2%) reported that they had been advised to express breastmilk during pregnancy, and 97 (14.2%) reported that they had undertaken breast expression during pregnancy (See Fig. 1).

Of the 97 participants who had expressed breastmilk during pregnancy, 38 had been advised to (39.2%), 57 (58.8%) had not been advised to and 2 (2.1%) could not remember if they had been advised to. Of the 56 women who reported that they had been advised to express breastmilk during pregnancy, 38 (67.9%) actually expressed, compared to 9.3% of women who had not been advised to express (Fig. 2).

Women who undertook ABE commenced expressing between 0 and 41 weeks of pregnancy, with a median of 36 weeks: 2 (2.2%) reported commencing ABE at 0 weeks of pregnancy, 5 (5.4%) between week 12 and 28, 43 (46.7%) between weeks 28 and 37, and 44 (45.7%) from 37 weeks onwards. Women who had practiced ABE reported giving birth at a mean of 39.7 weeks gestation in their last pregnancy, which was similar to the 39.3 weeks mean length of gestation reported by women who did not undertake ABE.

A majority of respondents, 398 (58.6%), were not sure if ABE was a good idea, with 233 (34.3%) of the respondents stating that they thought ABE was a good idea and 48 (7.1%) thinking ABE was not a good idea (See Table 2). Previous practice of ABE was statistically significantly associated with opinion of ABE (p < 0.001); the proportion of women who thought ABE was a good idea was highest in those who had undertaken ABE previously (78.4%), although 26.8% of those who had not previously undertaken ABE still thought ABE was a good idea.

A total of 547 (80.9%) of participants stated that they would consider doing ABE if it was found to help prepare for breastfeeding (See Table 2). Previous practice of ABE was significantly associated with whether participants would consider doing ABE in the future (p < 0.001), with only one participant (1.0%) who had previously done ABE reporting she would not consider doing ABE again if it was found to be helpful to prepare for breastfeeding.

In this sample, the proportion of women who had heard of ABE increased with increasing BMI group; 56.7% of women with a BMI of less than 25 kg/m² had heard of ABE compared to 63.2% of participants with a BMI within the overweight subgroup 78.7% within the obese subgroup (See Fig. 3). This association showed statistical significance (p < 0.001).

The proportion of women who thought ABE was a good idea also increased with increasing BMI group (See Fig. 3); 46.6% of obese participants thought it was a good idea, compared with 31.8% of those who were overweight and 30.2% of those with a BMI of less than 25 kg/m², and this association was statistically significant (p = 0.007). However, there was no significant association between BMI group and whether participants would consider doing ABE in the future (p = 0.280).

Several themes emerged from the qualitative data (See Table 3). Amongst participants who felt ABE was a good idea these were; beneficial when mother or baby have medical problems, and preparation for breastfeeding. The themes that emerged amongst participants who thought that ABE was not a good idea were; interfering with nature and harmful. For those who were unsure if ABE was a good idea the main theme was lack of knowledge.

Participants who thought that ABE was a good idea felt it would be beneficial where mother or baby have medical problems including gestational or pre-existing diabetes, previous breast surgery, complications after birth, prematurity, low blood sugars, admission to neonatal unit or difficulties feeding;

"If there are any complications during labour which meant that you were unable to feed initially (I.e. PPH [postpartum haemorrhage], or surgery was required taking you away from baby) baby could be spoon or cup feed colostrum. Equally if baby is struggling with blood sugars, jaundice, weight loss."

"I have type 1 diabetes, antenatal expression enabled me to collect colostrum and give it to my babies at birth preventing them from having low blood sugar."

They also referred to ABE as positive preparation for successful breastfeeding, including the opportunity to become more confident with the expressing technique that they could use again after the birth, establishing a supply of colostrum, encouraging the milk supply and avoiding the use of formula milk;

"A good idea to have some milk stored to avoid formula top ups if struggling to feed."

"I would have found it helpful…to have already got used to hand expressing as this was something I needed to do a lot once baby was born."

A few women also reported that ABE could be beneficial in promoting the onset of spontaneous labour;

"…It is used to induce labour naturally."

Participants who felt that ABE was a bad idea raised concerns that it would be harmful; causing preterm labour, painful to undertake, stressful, and result from bullying women to do it. They were also concerned that it was interfering with the natural process of initiating breastfeeding;

"I worried it would cause early labour."

"Nature gets this right, no need to interfere."

"Worried of making that [colostrum] go and turn straight into milk when baby arrives."

The predominant theme amongst the participants who were unsure if ABE was a good idea was their lack of knowledge; either that they had not heard of it or did not have enough information about the benefits or risks involved;

"Have never heard of it or its benefits /negatives."

"I didn't know it could be done."

"I…was never given any information."

The participants identified a number of factors that would encourage them to express breastmilk in pregnancy (See Table 4).

This survey included a large sample of women from a wide spread of UK locations. Comparing UK participants’ characteristics with national population data from England, demonstrated that this sample were more predominantly of a white ethnic group and there were a considerably lower proportion of Asian respondents. The sample was also older and of higher socioeconomic status, as indicated by occupation, than the current childbearing population, and hence some of their views and experiences may not be representative. The BMI distribution of the UK participants was very similar to that in the national population.

Undertaking this survey using online technology facilitated wide access and a high response rate that covered wide geographical areas with minimum resources. A maternity user group representative leading survey recruitment may have been an advantage to encourage such a large response rate. However, as a retrospective, self-reported questionnaire, it may have been subject to selective recall bias. The sample demonstrated a much higher breastfeeding rate than that of the national childbearing population in England; 98.4% of the participants reported they had breastfed, compared with the national breastfeeding rate of 74.3% at birth [52]. The participants had also breastfed for longer with 95.2% breastfeeding at 8 weeks compared to the national figure for England of 43.8% [52], and 84.0% reported breastfeeding at 6 months, compared to 34% of the national childbearing population [5]. This was therefore a self-selected sample of women who were highly motivated and successful breastfeeding mothers, and may not represent the opinions of the wider population. Nevertheless, even assuming a higher rate of motivation, a large proportion had concerns for implementing ABE and in depth analysis of their views are worthy of further investigation and consideration.