Babies born to women who have diabetes are known to have a higher chance of experiencing hypoglycaemia (or low blood sugar). They are more likely to be admitted to a special care baby unit and there is a lower chance that they will be exclusively breastfed, perhaps as a consequence of these other factors. So midwives and others have long encouraged women with diabetes to express their breastmilk in late pregnancy and store it, so that it can be used for their newborn if required. But, until now, we’ve not had enough evidence on whether this practice is effective. We also don’t know whether it is safe, and this is an important question because the act of expressing breastmilk involves the release of hormones which may cause uterine contractions. Previous research has suggested that it may cause birth to happen slightly earlier, but the quality of the research was low and the findings are not reliable (East et al 2014).
But the latest issue of The Lancet includes a paper describing a study which looked at these questions, and the answers suggest that, in healthy women who are at least 36 weeks pregnant, this practice is safe (Forster et al 2017). The research was a multicentre, two-group, unblinded, randomised controlled trial carried out at six hospitals in Victoria, Australia and a summary of the research is below. The results showed that no harm came from expressing and, while there was no difference in the chances of a baby being admitted to special care, there was moderate evidence that a higher proportion of the babies born to the first-time mums who expressed breastmilk received exclusive breastmilk during the first 24 hours of life. Interestingly, that association wasn’t there for the women who had given birth before, but the researchers pointed out that the study was really too small to have found this difference if it did exist, so we should not read too much into that. I found it especially interesting that there was evidence of mild uterine activity (as measured by CTG monitor) in the women who expressed breastmilk, but no episodes of tachysystole or hyperstimulation.
Research like this is hard to carry out, for a number of reasons that have much more to do with practicality and human factors than to the skill of the researchers. It is important to note that around 15% of the women allocated to the expressing group either did not express at all or did this less than five times, and it was impossible to blind some elements of this research. But this is still a really useful and interesting study which will hopefully reassure those women who want to be assured of the safety of this practice before they try it in the hope of helping their newborn baby.
East CE, Dolan WJ, Forster DA (2014). Antenatal breast milk expression by women with diabetes for improving infant outcomes. Cochrane Database Syst Rev 2014;7:CD010408.
Forster DA, Moorhead AM, Jacobs SE et al (2017). Advising women with diabetes in pregnancy to express breastmilk in late pregnancy (Diabetes and Antenatal Milk Expressing [DAME]): a multicentre, unblinded, randomised controlled trial 389(10085): 2204–2213. https://doi.org/10.1016/S0140-6736(17)31373-9
Background: Infants of women with diabetes in pregnancy are at increased risk of hypoglycaemia, admission to a neonatal intensive care unit (NICU), and not being exclusively breastfed. Many clinicians encourage women with diabetes in pregnancy to express and store breastmilk in late pregnancy, yet no evidence exists for this practice. We aimed to determine the safety and efficacy of antenatal expressing in women with diabetes in pregnancy.
Methods: We did a multicentre, two-group, unblinded, randomised controlled trial in six hospitals in Victoria, Australia. We recruited women with pre-existing or gestational diabetes in a singleton pregnancy from 34 to 37 weeks’ gestation and randomly assigned them (1:1) to either expressing breastmilk twice per day from 36 weeks’ gestation (antenatal expressing) or standard care (usual midwifery and obstetric care, supplemented by support from a diabetes educator). Randomisation was done with a computerised random number generator in blocks of size two and four, and was stratified by site, parity, and diabetes type. Investigators were masked to block size but masking of caregivers was not possible. The primary outcome was the proportion of infants admitted to the NICU. We did the analyses by intention to treat; the data were obtained and analysed masked to group allocation. This trial is registered with the Australian New Zealand Clinical Trials Registry, number ACTRN12611000217909.
Findings: Between June 6, 2011, and Oct 29, 2015, we recruited and randomly assigned 635 women: 319 to antenatal expressing and 316 to standard care. Three were not included in the primary analysis (one withdrawal from the standard care group, and one post-randomisation exclusion and one withdrawal from the antenatal expressing group). The proportion of infants admitted to the NICU did not differ between groups (46 [15%] of 317 assigned to antenatal expressing vs 44 [14%] of 315 assigned to standard care; adjusted relative risk 1·06, 95% CI 0·66 to 1·46). In the antenatal expressing group, the most common serious adverse event for infants was admission to the NICU for respiratory support (for three [<1%] of 317. In the standard care group, the most common serious adverse event for infants was moderate to severe encephalopathy with or without seizures (for three [<1%] of 315).
Interpretation: There is no harm in advising women with diabetes in pregnancy at low risk of complications to express breastmilk from 36 weeks’ gestation.