Transmission of SARS‐CoV‐2 through breast milk and breastfeeding: a living systematic review

Abstract

Abstract

The objective of this review is to assess available evidence related to the possible transmission of SARS‐ CoV‐2 through breast milk and through breastfeeding (i.e., related bodily fluids, such as blood, sweat, and respiratory droplets) or droplet transmission due to close contact with the infant or young child via skin‐to‐skin exposure or airborne transmission. The review aims to support policymakers in making evidence‐informed global, regional, and national guidelines on infant feeding in the context of the ongoing pandemic.

Introduction

On March 11, 2020, the World Health Organization (WHO) declared the coronavirus disease (COVID‐19) outbreak a pandemic, with more than 600,000 cases globally. 1 Infection with a novel coronavirus strain, SARS‐CoV‐2, causes COVID‐19. Clinical manifestation of SARS‐CoV‐2 infection varies; most infections are asymptomatic or present with mild symptoms, such as fatigue, fever, and cough. 2 Severe cases may progress to viral pneumonia and require oxygen therapy, intensive care, and mechanical ventilation. 3

There is limited but increasing evidence of SARS‐CoV‐2 infection during pregnancy. As initial reports on infections have been primarily from pregnant women confirmed or suspected with COVID‐19 who were infected in either the third or late second trimester of pregnancy, it is not yet entirely clear if pregnant women are at greater risk of becoming infected or if SARS‐CoV‐2 can be transmitted from mother to infant during pregnancy, delivery, or breastfeeding. 4 , 5 Similarly, information on COVID‐19 during infancy or in childhood remains sparse.

On May 27, 2020, WHO updated Interim Guidance on Clinical Management of COVID‐19, recommending exclusive breastfeeding for at least the first 6 months and breastfeeding alongside complementary foods until 2 years of age while using necessary precautions for infection prevention and control in infants born to mothers with suspected or confirmed COVID‐19 (WHO 2020a)^b. This recommendation is based on the health benefits associated with breastfeeding for both the mother and the child and the relatively mild or asymptomatic illness experienced by infants reported so far. Infant feeding practices are an area of concern for families and health personnel, particularly during such outbreaks and high uncertainty of associated risks, such as its impact on food and nutrition security.

The objective of this review is to assess available evidence related to the possible transmission of SARS‐CoV‐2 through breast milk and breastfeeding (i.e., related bodily fluids, such as blood, sweat, and respiratory droplets) or droplet transmission due to close contact with the infant or young child via skin‐to‐skin exposure or airborne transmission. The review aims to support policymakers in making evidence‐informed global, regional, and national guidelines on infant feeding in the context of the ongoing pandemic.

Methods

We designed and piloted a structured search strategy. The search was carried out on July 07, 2020, in the following electronic databases: MEDLINE (PubMed), the WHO COVID‐19 Global literature on coronavirus disease (https://search.bvsalud.org/global-literature-on-novel-co ronavirus‐2019‐ncov/), Cochrane Library, Web of Science Core Collection, and Embase (May 15, 2020). We also searched the COVID‐19 subset of the WHO International Clinical Trials Registry Platform (ICTRP) (July 07, 2020) to identify ongoing and unpublished studies. The WHO COVID‐19 global literature database is a comprehensive multilingual database on COVID‐19 updated daily (Monday through Friday) from searches of bibliographic databases, hand searching, and the addition of other expert‐referred scientific articles.

Living systematic review guidelines employ a continual approach to searching the literature on rapidly emerging research topics, to ensure greater currency and validity, and to increase the benefits to end users. 6 We used standard reporting methods described elsewhere. 7 Updated searches are planned to be performed as needed to keep the results up to date as the COVID‐19 research base grows over time. Keywords used in the search strategy will be updated every week, incorporating new terminology for the virus as it comes into use. A detailed presentation of the number of studies retrieved, deduplicated, excluded during screening, and included in findings as of the third iteration of the search update is given in Figure 1 (PRISMA flow diagram). The protocol for this review is registered in PROSPERO, an international prospective register of systematic reviews: CRD42020178664. 8

Open in a separate window

Figure 1

PRISMA chart (as of July 07, 2020).

Results

Our search initially identified 19,414 records, of which, 7926 titles and abstracts were screened after removing duplicates. A total of 605 full‐text articles were assessed for eligibility, of which 265 were excluded. A total of 340 reports were included, of which 37 included the analysis of breast milk samples and 303 described lactating women or infants and children 0–24 months of age without collection or SARS‐CoV‐2 testing of breast milk samples. There were 20 records where the full text could not be located and are thus awaiting assessment.

Among the 303 reports without breast milk sample collection, there were 73 included reports that did not contribute data to inform the results of this review due to lack of breast milk or outcomes of interest: 36 reports presented pooled data of children of breastfeeding age and older; 27 other reports included pregnancy outcomes but not neonatal or breast milk outcomes; in three reports, neonates were not tested; six studies described ongoing pregnancies; and two reported a spontaneous abortion or pregnancy termination. References to these studies can be found in Table 2.

Discussion

In this review, we assess the evidence on transmission of SARS‐CoV‐2 from the mother to her child through breast milk and breastfeeding (i.e., related bodily fluids, such as blood, sweat, and respiratory droplets) or droplet transmission due to close contact with the infant or young child via skin‐to‐skin exposure or airborne transmission. We also summarize outcomes for infants and children with suspected or confirmed SARS‐CoV‐2 infection according to breastfeeding practices, as reported by the authors.

Among the 37 included studies with breast milk samples, nine out of 84 analyzed breast milk samples were reported to be positive for SARS‐CoV‐2 RNA via RT‐PCR analysis 10 , 14 , 31 and one sample had specific IgG. 35 Among the cases with viral RNA detected in breast milk samples, one healthy neonate 10 had negative nasopharyngeal and anal test results for COVID‐19, while maternal breast milk samples had detectable viral RNA. However, this neonate was fed with a breast milk substitute, while maternal breast milk was positive and thus, it is not possible to ascertain the risk of infection by exposure to breast milk. In another case, the infant tested negative to the virus and exposure through breast milk could not be confirmed because infant feeding practices were not reported. 31 In a different case, a neonate was found to be positive for COVID‐19 based on viral RNA detected by RT‐PCR, while exposed to maternal breast milk that tested positive for SARS‐CoV‐2 RNA. 14 The mother breastfed the neonate while using surgical masks and following hygiene recommendations. In this particular case, before showing symptoms, both mother and newborn shared a hospital room with another mother–infant pair diagnosed with COVID‐19. Thus, it is not clear whether the newborn became exposed due to close contact with the other confirmed COVID‐19 patients, contact with the confirmed COVID‐19–positive mother, or through the exposure to her SARS‐CoV‐2 RNA–positive breast milk. Moreover, this newborn was also coinfected with respiratory syncytial virus and it is possible that the coinfection could have increased the neonate's vulnerability to COVID‐19 or worsen the symptoms of the infection. In the case of an 8‐month‐old infant with COVID‐19, 32 it is not possible to determine if transmission occurred through breast milk intake or breastfeeding, since it was interrupted during the maternal symptomatic phase, and the infant had symptoms 1 day after the mother's symptoms; additionally, mother and infant were staying in an area with ongoing COVID‐19 community transmission for 2 months.

Even though viral RNA has been detected in breast milk samples, among the included studies there were no attempts to culture the SARS‐CoV‐2 from breast milk isolates, adding to the uncertainty about potential infectious capacity of breast milk. One preprint study 49 found no culturable virus from one breast milk sample with viral RNA and from control breast milk samples spiked with SARS‐CoV‐2.

From the two reports that described breastfeeding practices, it is not possible to conclude if SARS‐CoV‐2 infection was due to mother‐to‐child transmission, which can include but is not limited to specific breastfeeding practices or to feeding young children with breast milk from a COVID‐19–positive woman. For the latter, the evidence gathered also includes another 79 breast milk samples, belonging to different women with confirmed COVID‐19, which had no traces of SARS‐CoV‐2 RNA, and this may suggest a low risk of transmission by ingesting breast milk. Interestingly, there were several confirmed COVID‐19 infants by RT‐PCR tests who received SARS‐CoV‐2 negative breast milk; these included two newborns exclusively breastfed, 21 one 27‐day‐old neonate who was breastfed, 15 and infants aged 3 and 6 months old. 17 , 25 Therefore, it seems plausible that these infants might have been exposed to SARS‐CoV‐2 through close contacts with infected family members, especially considering that in all these cases, both parents, relatives, and individuals in their communities were diagnosed with COVID‐19. Notably, infant feeding practices were scarcely reported in most of the included studies.

From the evidence reported in articles without breast milk samples tested, it is not possible to ascertain if there is an increased risk of viral infection among breastfeeding children via breast milk. Most breastfeeding neonates did not have evidence of COVID‐19 (28 out of 33) based on negative RT‐PCR tests. Similarly, among infants who were fed with a breast milk substitute, 63 out of 77 did not have COVID‐19 based on negative RT‐PCR tests. However, infant feeding practices were not accurately reported. From all these studies, the frequency or exclusivity of breastfeeding is unclear, especially in relation to onset of maternal symptoms or viral load.

Most neonates born to mothers diagnosed with COVID‐19 by RT‐PCR tests during their pregnancy were negative for viral infection. One neonate was reported to be negative for COVID‐19 by RT‐PCR tests in throat swabs but had elevated IgM and IgG levels at birth. 12 These findings may suggest that vertical transmission does not appear to occur during the peripartum period. It is possible that passive immunity from the mother to the infant could be protective against infection. This was also suggested in the context of severe acute respiratory syndrome coronavirus (SARS‐CoV) infection. 50 Given that all maternal infections reported in the included studies occurred shortly before delivery, the consequences of infection during earlier stages of pregnancy are not clear. Additionally, the majority of included studies did not assess the presence of SARS‐CoV‐2–specific antibodies in neonates, breast milk, placenta, or other tissues and hence, the significance of finding anti‐SARS‐CoV‐2 IgM immediately after birth in serum samples from the neonates, as reported by Ref. 12, remains to be defined. It is also important to understand the potential of passive immunity from mother to child during pregnancy and by breast milk intake. Specific IgA antibodies have been found in 12 of 15 breast milk samples, for example, from different women who had COVID‐19, 51 and further research is needed to ascertain the protective capacity and the duration of these antibodies.

It is also important to understand if the risk of transmission through breast milk changes in different stages of maternal disease progression or in asymptomatic cases. The three breast milk samples with detectable viral RNA were collected during the maternal symptomatic phase. 10 , 14 , 31 But the 43 breast milk samples without detectable viral RNA were collected during either the symptomatic or convalescent phase. This might suggest that SARS‐CoV‐2 does not always cross the alveolus or milk‐secreting unit during the acute phase, when maternal viremia is expected to be higher, or during the convalescent phase. Even after 4–5 weeks of maternal infection, viral RNA was not detected in breast milk, possibly indicating that the alveolus does not act as a reservoir for SARS‐CoV‐2.

Breast milk is not homogenous and its composition changes through the lactation period. In the future studies, it will be important to determine if the risk of transmission changes with the stage of lactation. In this review, most of the breast milk samples (43 out of 46) were negative for SARS‐CoV‐2 RNA presence, and the majority were tested during the first 48 h postpartum. This early period usually has a higher concentration of immune factors in milk compared with the composition of mature milk after the first month postpartum. There were few samples at different stages indicating the need to further characterize the risk of SARS‐CoV‐2 transmission in the context of breast milk composition, especially considering the immunomodulating components of breast milk, including antibodies, growth factors, and other proteins, with critical roles in sustaining healthy neonatal intestinal epithelium and with antimicrobial properties. 52 Lactoferrin, an iron‐binding protein present in breast milk, for example, has been found to inhibit SARS‐coronavirus infection in cell culture conditions. 53

What is known from SARS‐CoV

When we consider the past experience with SARS‐CoV, the viral agent that caused a major outbreak in 2003, it is important to note that it is genetically similar to SARS‐CoV‐2. 54 We identified two case reports where SARS‐CoV was tested in breast milk. 55 In one case, 56 the breast milk sample collected at birth was positive for SARS‐CoV. The infant was exclusively breastfed and did not show any symptoms of the disease at 5 days of age. The mother was diagnosed with SARS‐CoV infection at 20 gestational weeks, and she required intensive care. In a second case report, 55 the mother presented with serious symptoms of SARS‐CoV infection at 20 gestational weeks, and the infant was born healthy at 38 gestational weeks by cesarean section. Viral RNA was not detected in maternal samples of serum, whole blood, nasopharyngeal, and rectal swabs, or postdelivery, in the placenta, cord blood, amniotic fluid, and breast milk. However, antibodies to SARS‐CoV were detected in maternal serum, cord blood, and breast milk, suggesting that the newborn might have acquired passive immunity from the mother. Curiously, in both cases, the mothers developed gestational diabetes. 55 , 56

Another case series 57 reported five infants born to mothers with SARS‐CoV infection and pneumonia. All five infants tested negative for the presence of SARS‐CoV RNA in samples of serum, throat swab, and urine. They also had negative cultures, but serum samples indicated the presence of SARS‐CoV antibodies. The antibody titres did not increase in paired samples between the acute (1–9 days) and convalescent phases (21–23 days), suggesting passive immune transfer during pregnancy. 57 It is possible that the vertical transmission of SARS‐CoV may be prevented by the presence of antibodies in cord blood and breast milk, as previously described. 55 , 57 In summary, we reviewed the evidence for the transmission of the related SARS coronavirus (SARS‐CoV) and found that most studies reporting on infection status of mothers and their children did not describe breastfeeding behaviors or test breast milk samples. This evidence from the SARS‐CoV outbreak in 2003 can be useful in the interpretation of the results from this review on COVID‐19.

Limitations

Findings from this review are limited by the scarcity of information of infant feeding practices and breast milk samples tested, there was very limited information on skin‐to‐skin contact, rooming‐in, and exclusive breastfeeding. In most cases, there was not clear information on the breast milk sample collection and analysis procedure. We also identified several studies from diverse countries that reported infant infection outcomes, but the reporting of feeding practices was heterogeneous or not available. Additionally, there could be a high risk of publication and sample bias, given that the most severe cases are the ones that seek care, are tested, and most commonly only those with a positive diagnosis are most likely to be reported. Moreover, there could be duplication of reported cases as case reports become part of a larger body of evidence. We plan to update these review findings as the evidence continues to grow with upcoming results from cohort studies and other case reports.

Implications for future research

More evidence about the intake of breast milk directly from the breast or expressed, as well as infants being breastfed exclusively, or formula fed will further the understanding of SARS‐CoV‐2 transmission. Despite the rapidly increasing literature around SARS‐COV‐2 transmission routes, the evidence of possible transmission through breast milk is still limited. Most of the reported cases are symptomatic infants or symptomatic pregnant women and their newborns, the information on older infants with or without symptoms is scarce. In most cases, only one breast milk sample was tested and with no information on maternal viremia. Breast milk samples were only tested by RT‐PCR assays, and it is possible that viral RNA detection in breast milk was affected by the component of breast milk tested, as it has been shown to affect the assay sensitivity in other contexts. 58 It is important to note that viral RNA detection in breast milk does not necessarily indicate viral infectivity, and other assays will be necessary to determine if there are viral particles in the breast milk that can be infectious once ingested by children.

This systematic review included findings from 37 studies; 19 infants out of 77 were diagnosed with COVID‐19 by viral RNA detection and one by serology; 59 breast milk samples that were collected from COVID‐19–positive mothers tested negative for SARS‐CoV‐2 via RT‐PCR. As new evidence emerges from the current SARS‐CoV‐2 outbreak, additional research is warranted to explore the possible dynamic of breastfeeding transmission or protection against SARS‐CoV‐2. Specifically, studies focused on defining the timing of maternal and infant exposure, breast milk viral load, duration of infection, and the presence of protective antibodies in breast milk might aid in determining the risks of SARS‐CoV‐2 transmission when a breastfeeding woman is infected. We suggest that investigators working with breastfeeding mothers and their children collect the following data, where feasible, to better inform this review and the consequent guidelines: (1) confirmation and clinical characteristics of maternal and infant infection; (2) analyze and report on the presence of virus and viral infectious particles in breast milk, potentially with serial samples; and (3) record data on infant feeding practices and any contact precautions observed, such as wearing a mask and isolation.

The review authors are uncertain whether SARS‐CoV‐2 transmission via breast milk is possible as the certainty of the evidence has been assessed as very low. Nonetheless, the possible transmission through other breastfeeding‐related bodily fluids, such as blood, sweat, respiratory droplets, or droplet transmission due to close contact with the infant or young child via skin‐to‐skin exposure or airborne transmission cannot be disregarded.

Note adding in proof. Since this work was completed and accepted, Chambers et al. report that although SARS‐CoV‐2 RNA was detectable in 1 breast milk sample from 18 infected woman, this did not represent replication‐competent virus; the authors concluded that breast milk may not be a source of infection for infants. JAMA. Published online August 19, 2020.https://doi.org/10.1001/jama.2020.15580

Disclaimer

The authors alone are responsible for the views expressed in this review and they do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated.

Author contributions

S.M., P.R.S., J.L.F., and J.P.P.R. conceptualized the study. E.C.T., M.M.R., J.L.F., P.R.S., M.N.G.C., L.R., J.P.P.R., and S.M. designed the protocol. K.G.K. and E.C.T. designed, updated, and translated the search strategy. E.C.T., M.M.R., and P.R. screened and extracted data from articles. E.C.T., M.M.R., J.L.F., P.R.S., M.N.G.C., L.R., P.R., J.P.P.R., and S.M. interpreted the findings. E.C.T. and M.M.R. wrote the first draft, which all authors revised for critical content. E.C.T. and M.M.R. contributed equally to this study. All authors approved the final manuscript. S.M. is the guarantor. The corresponding authors attest that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding

This work was supported by the Division of Nutritional Sciences, Cornell University, and the World Health Organization, Geneva, Switzerland.

Supporting information

Acknowledgments

Footnotes

Contributor Information

Juan Pablo Peña‐Rosas, Email: tni.ohw@jsasoranep.

Saurabh Mehta, Email: ude.llenroc@athems.

References