January 7, 2025

expert reaction to study finding some maternal medications were associated with lower levels of protein and fat in breast milk

A study published in JAMA Network Open looks at maternal antidepressant and anti-Inflammatory medications and levels of nutrients in human breast milk. 

Dr Anna Gavine, a Senior Lecturer within the Mother and Infant Research Unit at University of Dundee’s School of Health Sciences, said:

“There is limited evidence on the safety of many medicines while breastfeeding so more research in this area is welcomed.

“Specifically, this paper looks at the levels of carbohydrates, fat, protein and energy human milk from mothers exposed to different medications: selective serotonin reuptake inhibitors (SSRIs), monoclonal antibodies (MAB), steroids and other anti-inflammatory drugs (ADs). 

“The milk from these women is compared to two control groups: women with untreated matched conditions and healthy women. Whilst it generally appears an appropriately conducted piece of research the authors do state there is uncertainty in the accuracy of the near-infrared milk analyser. However, I do have concerns about over-speculation and a negative focus in the reporting of some of the findings which could deter women from breastfeeding. Moreover, whilst some confounding factors were considered in the analyses (e.g. infant age parity, maternal BMI). There may be other factors which could have affected the results such as maternal diet.

“There was no statistically significant difference in carbohydrate or energy levels between the groups, which suggests that these medicines will not impact on the amount of energy a breastfed baby receives, and this finding is not sufficiently highlighted in the paper.

“There were some differences in fat and protein levels between some of the groups for some medications. However, these differences were small and generally not consistent and would suggest that women exposed to these medicines do not need to be concerned about the macronutrient content of their milk.

“More specifically, for women taking ADs there was a small but statistically significant difference in levels of fat, with women taking ADs having slightly lower levels of fat compared to disease matched controls. However, there was no difference in fat levels between women exposed to ADs and the healthy controls and again this is not sufficiently highlighted in the paper. Conversely, for women taking MAB, there was a small but statistically significant increase in levels of fat compared to the healthy controls. However, this did not remain after adjusting for confounding. Whilst this finding is reported in table 2, it is not discussed in the text of the paper.

“Protein levels were slightly lower in women taking MABs compared to healthy controls, but again this did not remain after adjusting for confounding. After adjusting for confounding, women on SSRIs and steroids had slightly lower levels of protein compared to healthy controls but not when compared to disease matched controls.

“Crucially, even with these potential differences the mean levels of all macronutrient ranges of milk from women in all groups in the study was within normal ranges. Therefore, these results should not change current breastfeeding policy and practice.

“Perhaps the most important medicine considered here is the SSRIs. Up to 10-15% of women experience more severe symptoms of post-natal depression that may require treatment with medications. However, concerns around medication can either inappropriately deter women from breastfeeding or lead to their mood disorder not being appropriately managed with medications, which has implications for the mother and baby relationship. The evidence that we do have suggests that sertraline (a SSRI) has been used in breastfeeding women without problems. It is therefore important that women requiring SSRIs are not inappropriately deterred from breastfeeding based on this research.”

Dr Edward Mullins, Clinical Senior Lecturer in Obstetrics, Imperial College London, said:

“I have concerns that the summary/press release for this study, ‘which finds some maternal medications were associated with lower levels of protein and fat in milk, which could impose health risks for breastfed infants. Other factors that could influence macronutrient levels need to be clarified before the clinical implications of these findings can be confirmed.’  may cause women who need medications for their health in pregnancy to stop taking them based on unfounded concerns, when the study actually showed that women taking medications in pregnancy had levels of breastmilk nutrients in the normal range. 

“The study did not compare like for like groups of women, comparing women with severe medical conditions requiring medication in pregnancy to those with less severe conditions not requiring medication. The severity of the women’s conditions and their dietary intake, which could affect breastmilk,  were not assessed- as the authors acknowledge. 

“In the comparison group not taking medication in pregnancy, those categorised as being likely to have mood disorders could have been over-estimated as the study used a low cutoff of 10 rather than 11-13 on the Edinburgh Postnatal Depression Scale¹ potentially leading to over-estimation of the number of women likely to have mood disorder in the group which did not take medication².

“The study risks being misleading because the summary of the conclusions, which is what most readers will see, says, ‘some maternal medications were associated with lower levels of protein and fat in milk, which could impose health risks for breastfed, infants. Other factors that could influence macronutrient levels need to be clarified before the clinical implications of these findings can be confirmed,’  i.e. there could be a concern. 

“…whereas the full conclusion says, ‘Low nutritional levels in human milk could negatively affect infant growth and impose health risks for the breastfed infant. However, the causality of these findings cannot be determined before the effects of other maternal exposures, such as maternal diet and the severity of the underlying disorder, are clarified. We conclude that as the mean macronutrient levels were within normal range in all exposure groups, these results should not affect the breastfeeding recommendations for mothers treated with antidepressants and anti-inflammatory medications, and the growth of their breastfed infants should be monitored as per clinical routine.’

“The underlined section would in my opinion have been a more appropriate summary conclusion.” 

1- https://med.stanford.edu/content/dam/sm/ppc/documents/DBP/EDPS_text_added.pdf

2- Levis B, Negeri Z, Sun Y, Benedetti A, Thombs B D. Accuracy of the Edinburgh Postnatal Depression Scale (EPDS) for screening to detect major depression among pregnant and postpartum women: systematic review and meta-analysis of individual participant data BMJ 2020; 371 :m4022 doi:10.1136/bmj.m4022

‘Macronutrients in Human Milk Exposed to Antidepressant and Anti-Inflammatory Medications’ by Heinonen et al. was published in JAMA Network Open at 16:00 UK time on Tuesday 7th January. 

DOI: :10.1001/jamanetworkopen.2024.53332

Declared interests

Dr Anna Gavine “Dr Gavine is a Senior Lecturer with research interests in women’s experiences of breastfeeding and how to more effectively support women to breastfeed. She receives no industry funding for her research.”

Dr Edward Mullins “No COIs relevant to this study.”