Breast milk and infection

Clin Perinatol 31 (2004) 501 – 528
Breast milk and infection
Robert M. Lawrence, MDa,*, Ruth A. Lawrence, MDb,c
aDivision of Pediatric Immunology and Infectious Diseases,
University of Florida College of Medicine, Health Science Center, 1600 SW Archer Road,
R1-118, Gainesville, FL 32610-0296, USA
bDivision of Neonatology, Departments of Pediatrics and Obstetrics and Gynecology,
University of Rochester School of Medicine and Dentistry, Rochester, NY, USA
cThe Breastfeeding and Human Lactation Study Center, Strong Memorial Hospital,
601 Elmwood Avenue, Rochester, NY 14642, USA
The nutritional, cognitive, emotional, and immunologic benefits of human
breast milk and breastfeeding are significant and well documented [1]. Human
milk protects against specific pathogens (viruses, bacteria, and parasites) as well as
separate clinical illnesses (eg, necrotizing enterocolitis, bacteremia, meningitis,
respiratory tract illness, diarrheal disease, and otitis media) [2]. The benefits of
breast milk make it one of the most important factors in protecting infants against
the morbidity and mortality of infectious diseases [2 – 8]. Numerous factors within
human breast milk act in a complementary fashion to protect against infection.
They contribute to the infant’s immune protection through various mechanisms,
including improved growth of nonpathogenic flora, decreased colonization with
enteropathogens, enhanced development of the respiratory and intestinal mucosal
barriers, specific factors against individual organisms (eg, secretory IgA [sIgA]),
functioning immune cells (eg, neutrophils, macrophages, T and B lymphocytes),
decreased inflammatory reaction, and immunomodulation [2,9 – 11].
Microorganisms also have been identified in colostrum and breast milk. Few
are readily transmitted through breast milk to cause clinically significant in-
fections in infants and children (eg, HIV1, human T-lymphotrophic virus I
[HTLV-I]). Others have been reported to cause infection in the infant rarely, af-
ter transmission through breast milk (eg, group B streptococci). Any decision
about possible infection of an infant or child through breast milk should weigh
the tremendous benefits of breastfeeding against the potential risk for transmis-
sion and the possible severity of the illness.
This article presents an overview of the considerations for breast milk and
infection, focusing on the most important organisms that are transmitted through
* Corresponding author.
E-mail address: lawrerm@peds.ufl.edu (R.M. Lawrence).
0095-5108/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.clp.2004.03.019

---

502
R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
breast milk (HIV1, HTLV-I, and cytomegalovirus [CMV]) and touching on other
organisms that are important in neonates and infants or that have captured public
notice. The basic assumption is that breastfeeding is contraindicated rarely during
maternal infection. The few exceptions are specific organisms with clear evidence
of transmission through breast milk that cause significant morbidity and mortality
because of infection through breast milk.
Essential concepts
Several factors must be considered to prove that breast milk is the mechanism
of transmission for a clinically significant infection in the infant (Box 1).
Excluding other probable mechanisms of transmission can be challenging,
especially in the neonatal period. The organisms that are a concern for trans-
mission through breastfeeding are transmitted more commonly prenatally, peri-
natally, and postnatally through other mechanisms. Congenital infection can
occur in any trimester, depending on the organism. The timing of infection also
significantly affects the clinical course of the infection in the fetus or newborn
Box 1. Factors to prove infection transmission through breast milk
1. Identify the infectious agent in the colostrum or breast milk
using culture, nucleic-acid detection, antigen detection, or
other methods.
2. Postulate a reasonable mechanism of infection given the
current knowledge about the organism and the host’s
response to it, including timing and pathophysiology.
3. Characterize the clinical manifestations of the illness in the
mother and the infant.
4. Demonstrate the occurrence of infection in the mother and
the infant by culture, nucleic-acid identification, or immuno-
logic response.
5. Confirm that the organisms identified in the mother and infant
are identical based on serotype, sensitivity pattern,
nucleic-acid sequence, and other factors.
6. Document that the risk for infection is greater in breast-fed
infants than in formula-fed infants.
7. Characterize a dose – response relationship between the
quantity of organisms in the breast milk as well as the amount
of breast milk ingested and the frequency of transmission or
the severity of the infection in the infant.
8. Exclude other possible mechanisms of transmission.
9. Prove the transmission through breast milk by reproducing
the process in animal or human studies.

---

R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
503
(eg, asymptomatic, fetal demise, prematurity, clinical disease present at birth, or
late presentation). Perinatal infections commonly occur from exposure to blood
or body fluids and contact with pathogens from the maternal genitourinary and
gastrointestinal tracts. The likelihood that the exposed infant will be infected
varies significantly with the specific organism and various host factors (eg,
passively acquired antibody levels in the infant). Postnatally acquired infections
are transmitted most commonly through contact with caregivers (eg, parents,
relatives, visitors, health care providers), the environment (eg, medical equip-
ment, other fomites), or breast milk, depending on the organism [12].
The timing of the infection in the mother and the infant is often crucial to
documenting the mechanism of transmission. In most infectious situations, the
exposure of the infant and mother to child transmission has occurred before the
illness is diagnosed in the mother (eg, measles, Coxsackievirus infection) and
frequently occurs before the mother becomes ill (eg, chickenpox, hepatitis).
Proscribing breastfeeding at that point may not prevent infection in the infant and
will diminish significantly the effect of breast milk to limit or modify the illness
in the infant. The timing of the infection in the mother can influence significantly
the risks to the infant (eg, primary HIV or CMV infection before pregnancy,
during pregnancy, or postnatally, ie, during lactation).
The approach to any suspected infection in the breastfeeding mother – infant
dyad should be systematic (Box 2). The mother’s initial immunologic response to
Box 2. Approach to a suspected infection in a breastfeeding mother
1. Suspect particular infectious agents based on the
clinical presentation.
2. Initiate a directed diagnostic work-up to identify the etiologic
cause of the infection.
3. Consider the probable mechanisms of transmission, the
known virulence of the likely infectious agents, and the
susceptibility of the infant.
4. Institute preliminary infection-control precautions (including
temporarily holding or continuing breastfeeding) based on the
clinical syndrome, site of infection, probable mode of trans-
mission, and infant’s susceptibility.
5. Start empiric therapy in the mother as indicated by the
severity of the mother’s illness.
6. Consider preventive or empiric therapy for the infant when
the risks for infection and significant disease are high.
7. Modify the empiric therapies and precautions based on the
agent identified as the cause of the infection in the mother.
8. Observe the mother for response to therapy and the infant for
signs and symptoms of infection requiring treatment.

---

504
R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
the infection may add specific factors to the breast milk that can prevent infection
or ameliorate the illness. Many maternal illnesses associated with fever do not
require separation of the mother and infant (eg, engorgement of the breasts,
atelectasis, nonsuppurative phlebitis, or urinary tract infection) or additional pre-
cautions to protect the infant.
Another important consideration relative to breast milk and infection is
medications in breast milk. A thorough review of antimicrobial agents and breast
milk is outside the scope of this article. Most antimicrobial agents used to treat
infection can be used in infants and children. Additional amounts that are ingested
by the infant in breast milk are usually insignificant compared with doses used to
treat the infant. In almost all cases, an antimicrobial agent that is appropriate for
treating the mother and compatible with breastfeeding can be selected.
Infection-control issues
Infection-control guidelines are intended for hospitals, but the principles of
epidemiology can be applied to any infectious situation, even in the home, and
can facilitate logical and reasonable use of interventions to prevent transmission
of infection to an infant.
Newer concepts and terminology have been proposed recently [13]. Standard
precautions include avoiding direct contact with blood and body fluids, nonintact
skin, and mucous membranes of every patient, regardless of the patient’s diag-
nosis, based on the idea that infection can be transmitted without an identified
infection and from unidentified sources. Standard precautions emphasize careful
handwashing with every patient contact and the use of appropriate barriers when
contamination with body fluids is more likely (eg, gloves, masks, glasses). Breast
milk is not considered a potentially infectious body fluid under these guidelines.
Careful handwashing before and after breastfeeding is always appropriate, as are
selected barriers (mask, gown or clothes, bandages) to prevent infant contact with
other body fluids of the mother with a specific infection. Washing the breast
before or after breastfeeding is unnecessary and may irritate the breast.
Specific contact, droplet, or airborne precautions are used for certain illnesses
and microorganisms based on the predominant mode of transmission [14].
Airborne precautions are used to prevent transmission through droplet nuclei
(ie, respiratory particles less than 5 mm in diameter that can contain microor-
ganisms). Respiratory protective devices (requiring personal fitting and seal
testing) are recommended for illnesses such as measles, varicella, disseminated
zoster, and tuberculosis. Because infants cannot wear such devices, infants of
mothers with these infections should be separated temporarily from the mother
during the infectious period, regardless of the mode of feeding. The infant can
receive the mother’s expressed breast milk through a bottle (given by another
individual), except when there are lesions of varicella-zoster or tuberculosis on
the breast, and in the case of measles, where breast milk should be held until the
infant has received immunoglobulin.

---

R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
505
Droplet precautions are used for larger respiratory droplets that travel only
short distances in the air and are transmitted more frequently by direct contact with
mucous membranes, hands, or objects contaminated with respiratory secretions.
Use of a surgical mask (by the individual with the infection) to limit the spray of
droplets and touching of mucous membranes as well as careful handwashing are
the primary interventions. Droplet precautions should be used for adenovirus,
diphtheria, influenza, Haemophilus spp, mumps, mycoplasma, Neisseria spp,
pertussis, respiratory illnesses, rubella, and Streptococcus spp. Timing is crucial
relative to the institution of such precautions to prevent disease. In the case of
parvovirus, a previously well mother is unlikely to be infectious after onset of the
rash [15]. Expressed breast milk can be given to the infant in most cases.
Contact precautions are used to prevent transmission by direct or indirect
contact with potentially pathogenic organisms. Indicated interventions can in-
clude the use of gloves and gowns, handwashing before and after using such
barriers, and cohorting, separation, or a private room. Contact precautions can be
used for many organisms (eg, diarrheal agents, multi-drug – resistant organisms).
The use of expressed breast milk is acceptable in most situations where contact
precautions are recommended, with Ebola virus infection or Lassa fever as no-
table exceptions [16].
Bacterial infections
Bacterial infections in neonates and infants are common. Systemic bacte-
rial infections occur in the neonate with a frequency of one to five episodes in
1000 live births. The timing of infection in the neonate is divided commonly into
early-onset (before 7 days of age, especially less than 24 hours of age), late-onset
(7 – 30 days of age), and very late – onset (after 30 days of age). The predominant
organisms of early-onset are group B streptococci and enteric bacilli, especially
Escherichia coli. Less common early-onset pathogens include other streptococci,
Enterococcus spp, Listeria spp, Haemophilus influenzae, Streptococcus pneumo-
niae, Chlamydia spp, and other organisms in the maternal genital flora. These
organisms can cause late- or very late – onset bacterial infections. Additional gram-
negative bacilli are resistant frequently to antibiotics, and Staphylococcus aureus
and coagulase-negative strains become more frequent, especially during hospital-
ization of a neonate or infant. Transmission of these organisms from mother to
child through breast milk is relatively rare compared to the transmission peri-
natally during delivery, or through direct contact with the mother, family, or health
care providers after birth. Table 1 lists selected organisms.
Botulism most frequently occurs between 6 weeks and 6 months of age and
primarily before 12 months of age, with the youngest patient in the literature
6 days of age [17]. Botulism is caused by the neurotoxin produced by Clostridium
botulinum. Arnon and colleagues [18] reviewed 50 patients hospitalized for botu-
lism in California and the breast-fed infants were older at diagnosis than formula-
fed infants, and had milder disease. In cases of sudden infant death syndrome

---

506
R.M.
Lawr
ence,
R.A.
Table 1
Lawr
Breastfeeding issues for selected bacterial maternal infections
Predominant modes
Evidence for transmission
ence
Organisma
of transmissionb
Usual timing of infectionc
in breast milk
Clinical significanced
/
Clostridium botulinum
Food-borne
NA
None
BF/BM
Clin
Toxin-mediated disease
Perinatol
Chlamydia trachomatis
Contact-genital, secretions
Perinatal
None
BF/BM
Escherichia coli
Contact—GI tract or stool
Perinatal
None
BF/BM
Haemophilus influenzae
Contact
NA
None
Delay BF 24-hr therapy
31
in mother; BM Rifampin
(2004)
prophylaxis for the infant.
Droplets
Listeria monocytogenes
Contact—GI tract
Perinatal
None
BF/BM
501–528
Food-borne
Postnatal
Mycobacterium tuberculosis
Airborne
Postpartum (perinatal and
Only with TB
With active TB, delay BF
congenital are rare)
Mastitis and lesions
for 14 days; maternal therapy
on the breast
or PI, BM.
Droplet
Infant—isoniazid prophylaxis.
Neisseriae gonorrheae
Contact—genital, oral, rectal
Perinatal (postnatal is rare)
None
Ceftriaxone—no delay, BF / BM.
Body fluids
Other medications—delay 24-hr
maternal therapy, BM.

---

Staphylococcus aureus
Contact
Postnatal
Lesions on the breast
Delay 24-hr maternal therapy, BM.
or mastitis
Contaminated, stored
Avoid BM with breast lesions
breast milk
or MRSA.
Coagulase- negative staphylococci
Contact
Postnatal (premature, sick
None
BF/BM
R.M.
neonates; IV lines, antibiotics)
Group B streptococci
Contact—genital tract,
Prenatal
Case reports, rare
Preventive therapy.
Lawr
secretions
Perinatal
Delay 24-hr maternal therapy, BM.
Postnatal
Empiric prescription for infante
ence,
This is a selected, limited list intended to consider some important bacteria that cause infection in the neonate or infant and possible issues related to breastfeeding and
R.A.
breast milk.
Abbreviations: BF, breastfeeding is appropriate; BM, expressed breast milk is appropriate; GI, gastrointestinal; IV, intravenous; NA, Not applicable; PI, period of
Lawr
infectivity; TB, tuberculosis.
a
ence
Bacteria that cause various clinical illnesses in the infant and the mother; the specific illnesses are too numerous to list.
b For breastfeeding and non-breastfeeding situations; does not include all possible or reported modes of transmission (airborne, body fluids, contact, droplet, food-borne).
/
c Does not include all possible times of transmission. If ‘‘NA’’ the timing of infection is not associated frequently with pregnancy, delivery, or neonates and infants.
Clin
d Notes the appropriateness of breastfeeding or use of breast milk when the mother has specific bacterial infection.
Perinatol
e Refer to the text for explanation.
31
(2004)
501–528
507

---

508
R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
associated with botulism, no infants were breast-fed within 10 weeks of death. An
association with the introduction of solid foods has been suggested. Honey and
corn syrup often are implicated because they may contain C botulinum spores. The
apparent immunologic benefit of breast milk may be caused by more acidic stools
(pH 5.1 – 5.4) and increased Bifidobacterium species in the stools of breast-fed
infants, limiting the presence of C botulinum or spores. Botulinum toxin pro-
duction declines with lower pH. No evidence suggests that the organism or the
toxin is transmitted through breast milk.
Chlamydia infection may be the most frequent sexually transmitted disease in
the United States. Perinatal infection in the infant produces conjunctivitis and
pneumonitis and is caused primarily by colonization of the infant when passing
through the birth canal. Specific sIgA has been identified in colostrum and breast
milk. There is no evidence for transmission through breast milk.
Escherichia coli is a common cause of neonatal systemic bacterial infection as
well as urinary tract infections and bacteremia in infants. E coli is ubiquitous in
the mother and infant’s environment. Breast milk has never been documented as a
source of E coli infection.
Haemophilus influenzae infections have decreased significantly in countries
with widespread use of the H influenzae conjugated vaccines (HibTITER,
PedvaxHIB, Comvax, ActHIB, OmniHIB) [19]. Transmission is through direct
contact and respiratory droplets; there is no evidence for its transmission through
breast milk. Breast milk seems to limit colonization of H influenzae in the infant’s
throat [20]. In the unusual situation of infection in a breastfeeding mother with
an incompletely immunized infant, then chemoprophylaxis is indicated for all
household members, including the infant [21]. Temporary separation of the infant
and mother is appropriate during the first 24 hours of the mother’s antimicrobial
therapy, after which breastfeeding can resume. Expressed breast milk can be
given to the infant in the interim.
Listeriosis during pregnancy infrequently causes premature delivery or still-
birth. Perinatal infection is an uncommon cause of severe disease in the neonate
with transmission through transplacental spread, infected amniotic fluid, or con-
tact with the organism in the maternal genital tract. No published information
suggests transmission of Listeria monocytogenes through breast milk. Breast-
feeding or use of expressed breast milk from the mother with Listeria infection is
appropriate, including the selection of antimicrobial agents that are compatible
with breastfeeding to treat the mother.
Neisseria gonorrhoeae is transmitted during passage through the birth canal
and infrequently from postnatal contact with the mother or her partner. There is
no documented risk for transmission in breast milk. Breastfeeding can continue
when the mother is treated with ceftriaxone (Rocephin), but a temporary ces-
sation (first 24 hours of maternal therapy) of breastfeeding and breast milk should
occur when other antibiotics are used.
Staphylococcal infection usually occurs late in the neonatal period. Forty to
ninety percent of infants in the nursery at 5 days of age will be colonized with
Staphylococcus aureus [22]. Staphylococcus aureus caused nursery outbreaks in

---

R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
509
the past. Postnatal contact with mothers, health care workers, and contaminated,
unpasteurized, banked breast milk were the identified sources of infection [23].
Staphylococcus aureus is a common cause of mastitis in the mother. There is one
reported case of staphylococcal scalded skin syndrome in an infant whose mother
had a skin lesion caused by staphylococcus on her areola [24]. No attempt to
identify the toxin or the organism in the breast milk was made. This case reinforces
close, continued observation of the infant (breast- or formula-fed) when there is a
documented maternal infection. In a case of toxic shock syndrome in a mother at
22 hours postpartum, the breast-fed infant remained well through 60 days of age.
Staphylococcal enterotoxin F (SEF) was identified in breast milk on days 5, 8,
and 11, but Staphylococcus aureus was isolated only from the mother’s vagina and
not the breast milk [25]. SEF is inactivated by pepsin at pH 4.5. It is probably
inactivated in the stomach, presenting no risk to the infant. Methicillin-resistant
Staphylococcus aureus (MRSA) is more frequent than in the past. Skin and nares
are the predominant sites of colonization. No treatment regimen to eradicate colo-
nization has proved highly successful. Current regimens include oral systemic
therapy with one or two sensitive antibiotics, topical antibiotics twice daily to the
nares, and intermittent bathing with hexachlorophene (pHisoHex) or a similar
agent. In the face of maternal Staphylococcus aureus infection, use of expressed
breast milk is appropriate during a temporary 24-hour separation of mother and
infant at the initiation of maternal antimicrobial therapy.
Coagulase-negative staphylococcal infection causes late-onset disease in sus-
ceptible neonates. Factors associated with increased risk for this infection include
prematurity, low birth weight, very low birth weight, invasive therapies (eg, in-
travenous lines, chest tubes, surgery, dialysis), antibiotic use, and prolonged
hospitalization. Colonization rates are as high as 60% to 90% for infants hos-
pitalized at 2 weeks of age, in selected nurseries. There is no difference in the
infection or colonization rates for formula-fed and human milk – fed infants.
Breast milk can be given and may provide significant other benefits to these
susceptible infants.
Group B streptococcus (GBS, Streptococcus agalactiae) is transmitted pri-
marily in utero and during delivery. The revised guidelines proposed by the
American Academy of Pediatrics committees on Infectious Diseases and the
Fetus and Newborn use several variables to identify increased risk for GBS in-
fection in the neonate and recommend intrapartum prophylaxis for those infants
at high risk [26]. Colonization of the infant during the postnatal period occurs
[27,28]. Although many infants are colonized, few develop disease [29]. Ac-
quisition of GBS through breastfeeding or breast milk is rare, but has been
documented in cases of late-onset GBS disease [30,31]. Butter and DeMoor [32]
demonstrated GBS in the nose and throat of infants when GBS also was cultured
from the mother’s breast. It is more likely that transmission occurs through con-
tact rather than the organism passing in the breast milk. A mother or infant colo-
nized or infected with GBS should be managed with standard precautions
during hospitalization. Routine culturing of the breast or breast milk and therapy
to eradicate colonization have not proved useful. GBS has not been associated

---

510
R.M. Lawrence, R.A. Lawrence / Clin Perinatol 31 (2004) 501–528
with outbreaks in the nursery. When the mother begins treatment for GBS disease
(most often endometritis), temporary, 24-hour separation of the mother and infant
should occur with the provision of expressed breast milk for the infant.
Tuberculosis (TB) is uncommon in the United States, but remains a significant
and common disease worldwide. Congenital TB is extremely rare, with fewer than
300 reported cases in the literature. TB mastitis is also rare. Starke [33] has
summarized the evaluation and treatment of a pregnant woman with a positive
tuberculin skin test. The primary concern is the postnatal exposure of the infant
through droplets or droplet nuclei by the mother or another household member
with active pulmonary TB. Complete evaluation of the suspected person to
determine their TB status (active disease with or without positive cultures and
smears) and testing of all household contacts are the first steps. Breast-fed and
formula-fed infants are equally at risk from respiratory transmission. Separation of
the infant from any case of active pulmonary TB is appropriate. Once adequate
therapy in the mother has begun and the mother is determined to not be infectious,
the infant and mother may have contact. Observation of the mother and infant
should continue through the completion of treatment for the mother. Transmission
of TB in breast milk has never been documented in the absence of TB mastitis.
Expressed breast milk can be given safely to the infant because antituberculous
medications can be used in infants. The only contraindication to using breast milk
is in the mother who has TB mastitis. Prophylactic isoniazid (Nydrazid, Laniazid)
therapy for the infant prevents TB infection in infants. Once both the mother and
infant are being treated and closely observed, they can be in contact [34].
Viral infections
Most significant viral infections in neonates or infants occur through trans-
placental or intrapartum transmission. The risk for transmission from mother to
child varies significantly if the maternal infection is a primary infection (eg,
herpes simplex virus [HSV], HIV1), a secondary (reactivation) infection (eg,
HSV, CMV) or a chronic infection (eg, hepatitis B, HIV1, HTLV-I) during
pregnancy or lactation. The transmission of infection through breast milk is well
documented for CMV, HIV1, and HTLV-I. Exposure to small amounts of virus in
human milk multiple times a day over the period of breastfeeding (months to
years) probably contributes to the high rate of transmission of CMV, HIV1, and
HTLV-I through breast milk. For most other viruses, transmission through breast
milk is rare (Table 2).
CMV is the most common cause of congenital infection in the United States.
Approximately 1% of all infants excrete CMV in their urine at or soon after birth
(less than 3 weeks of age). About 5% of the CMV congenitally infected infants
will manifest disease at birth and 15% will manifest congenital infection later (eg,
progressive late-onset hearing loss, learning disability) [35]. Perinatal infection
occurs through direct contact or body fluid contact at delivery, but is associated
rarely with clinical illness in full-term infants. Postnatal infection occurs through

---

Document Outline

• Breast milk and infection
  • Essential concepts
    • Factors to prove infection transmission through breast milk
    • Approach to a suspected infection in a breastfeeding mother
  • Infection-control issues
  • Bacterial infections
  • Viral infections
  • Other infections
  • Potential infections of bioterrorism
  • Emerging infections
  • Summary
  • Acknowledgements
  • References

---