Evidence on antenatal breast milk expression is very limited as few studies[5, 7, 15] have ever been published on this topic. In our study, less than half of those women who recalled being advised to express breast milk complied with this recommendation and reasons for this need to be explored.
The findings on gestational age and admission to SCBU were clinically important as there were indications of differences between the two groups. Considering 95% confidence intervals, we would need a larger sample size to estimate the true differences with more precision. Our results of slightly lower gestational age were in line with the findings from Kavanagh et al., which showed a link between breast stimulation and induction of labour. Forster et al. also found that more babies from the breast expression group were admitted to SCBU. They hypothesised that this finding could be by chance or could be due to the mothers’ reluctance to allow formula feeding, as the majority of admissions were due to hypoglycaemia. It may also be due to a lower gestational age at birth, as indicated in our sample, induced by breast stimulation in mothers who had expressed antenatally.
All the women in the antenatal expression group breastfed as their first feed, with perhaps the antenatal expression giving those women confidence and practice so they were psychologically and physically prepared to breastfeed after the birth. However, it could be simply because of their higher rate of motivation and interest in breastfeeding. Eighty-five percent of women with diabetes who did not express antenatally also initiated breastfeeding. This, in line with findings from Simmons et al., is of importance, as it may indicate a high level of commitment in initiating breastfeeding among women with diabetes in spite of the possibility of having delayed lactogenesis. This is particularly worthy of notice as the breastfeeding initiation rate in local women without diabetes was 71% at birth.
There is a strong association between obesity and gestational diabetes, with obese women less likely to initiate breastfeeding and more likely to stop early. Lower socio-economic status has also been shown to be an important predictor of lower breastfeeding rates in the general population and in women with diabetes. Although a larger proportion of women with a higher level of education (38%) expressed antenatally compared to 13% of women who left education by 19 years, the differences among the levels of education were not statistically significant between the two groups (Table1). The lack of significant differences (as an indicator of socio-economic status) may be due to the limitations with the sample size.
The trends identified in our study, although small in sample size, can inform the design of larger projects and raise attention to the importance of confounding factors such as co-morbidities and psychosocial factors.
The reasons why women chose to follow or ignore the advice to perform antenatal hand breast milk expression were not explored in this study. Further in-depth research on communicating such messages and the underlying decision-making processes by women for antenatal breastfeeding expression, with a particular focus on issues related to acceptability of the practice, women's concerns and cultural background, would be beneficial.
This study was retrospective and self-reported questionnaires are prone to selective recall bias, although information such as birth weight, gestational age and specific complications were checked in maternal records which took precedent over recall. The average time postpartum that participants received their questionnaire was 20 months, which was a considerable time to remember details and memories could be confused with any subsequent pregnancies. However, experiences in pregnancy, birth and the postpartum period are significant events in a woman's life, so accurate recall is likely to be fairly high.
The sample size was fairly small, and from one hospital in the UK, so may not be the same in other parts of the country, or with other populations of women who may have different attitudes and experiences of antenatal breast milk expression. The small sample size may also have contributed to the lack of statistically significant differences between the groups. The study response rate was 40% and so it could be presumed that those most engaged would have returned their questionnaires and taken part, however the general characteristics of the subjects were similar to those from national data.