There is growing interest in the postpartum period. It is a time of healing, repair and asymptotic achievement of pre-pregnancy states in many organ systems. The postpartum is “the fourth trimester” when mechanisms of poor pregnancy outcomes can be revealed or examined, and factors engaged in the beginning of pregnancy may be observed as “the system” returns to homeostatic set points. It is also when future risk assessment can be initiated. In addition, it has been hypothesized that the postpartum period may be a unique period for intervention against ongoing and subsequent risk. In particular, behavioral interventions, e.g., diet, exercise, and counseling have been supported as means to ameliorate future risk². One behavioral intervention which has long been supported is that of breastfeeding.

There is significant evidence that breastfeeding has many benefits to newborns, both nutritionally and immunologically³–⁵. However, evidence also suggests that breastfeeding exerts positive effects on the mother’s overall emotional⁶,⁷ and physiological health including those on lipid and glucose metabolism⁸, and on tissue specific and systemic vascular biology⁹–¹³. Data on healthy mothers that thoroughly investigates the postpartum and breastfeeding period is evolving. Existing studies suggest that breastfeeding may contribute to the observed persistence of pregnancy-related adaptations which, in the case of the cardiovascular system, may persist up to one year after birth¹⁴. The idea that breastfeeding may prolong pregnancy-related immunological adaptation has been suggested, but not fully investigated¹⁵. A clear picture in this area is perhaps hampered by both a lack of experimental and observational data, but also a classically limited view of maternal immunity¹⁶. This review is intended as a preliminary exploration of the potential role and importance of breastfeeding in postpartum immune homeostasis.

Background

THE IMMUNE SYSTEM AND THE DEVELOPING BREAST

Some have wondered whether the proteins expressed in the developing breast during puberty or pregnancy might be new to the immune system and therefore a possible target of the immune response¹⁷,¹⁸. To some this might only be true if there was an inherent defect or dysregulation in the development processes of white adipose tissue expansion, neuronal network development, duct epithelial cell turnover and cyclic micro-alveolarization. Evidence suggests that there might be a unique population of immune cells that participate in this process, including eosinophils¹⁹ and macrophages²⁰ which may participate in clearance of apoptotic debris, modulation of extracellular matrix components, angiogenesis or in important cell-cell signaling pathways. The high expression of TLRs on macrophages in the developing breast make them efficient at inciting an inflammatory response in the face of “opportunistic” infection with local microbes, but this inflammatory response can enhance any dysregulation in breast tissue components. T cells are also present within breast ducts, and data suggests that they may regulate epithelial cell proliferation and duct outgrowth and branching²¹. This regulatory function may depend on activation by local dendritic cells which presumably may present ductal peptide antigens²². This raises the interesting possibility of a unique, developmentally programmed regulatory T cell pool, as has been suggested for other tissues²³. One could also hypothesize that populations of regulatory B cells, for example capable of removing “self-antigens” generated by cellular homeostasis²⁴, may also participate in early breast development. These cell populations undergo their own fluctuations due to trafficking, death, proliferation and change of phenotype during pregnancy and lactation.

LOCAL IMMUNITY IN THE INVOLUTING BREAST

Post weaning, breast alveolar epithelial cells undergo apoptosis and clearance, and adipocytes return to their pre-pregnancy state. Any residual milk products must also be removed. Dysregulation or incompleteness of this process has been associated with defects in the populations of tissue resident macrophages (reviewed in²⁰). Macrophages in the involuting breast show an increasingly “alternatively activated phenotype”, including expression of arginase and the mannose receptor, and expression of molecules such as IL-10, which may be “suppressive” but may also contribute to regulation of angiogenesis²⁹ and wound healing³⁰. A population of macrophages in the involuting breast express an immature phenotype, but it is not clear is these are new immigrants as opposed to tissue resident cells²⁰,³¹.

Other immune cell subsets experience regulated presence in involuting versus lactating mammary tissue. Dendritic cells in mice stay relatively constant through lactation, increasing at the time of involution and peaking after weaning³². They also experience phenotypic changes such as the reduction of molecules such as CD80 and CD86 until weaning. Evidence suggests however that these cells may have a lower capacity to activate naïve T cells. CD4 T cells also increase at involution and weaning. The CD4 T cell population expresses an activated and mixed phenotype of Th-17, Th2 and Treg like cells that evolved through involution. During involution, there may be an accumulation of memory CD4 T cells³². Emerging data also suggest the importance of B cells in the involution process³³. The regulation of these cell types is very important in modulation of tumorigenesis (see below).

SEX HORMONES AS IMMUNOMODULATORY MOLECULES

Lactation produces lower circulating estrogen and progesterone in humans³⁴ and animals³⁵,³⁶.

The exact underlying mechanisms are still under study, but likely relates to dysregulated production of luteinizing hormone or gonadotropin releasing hormone and lack of ovulation. Both estrogen and progesterone have immunomodulatory capacity, both in the development and function of immune cells (recently reviewed in³⁷). Estrogen signaling has a complex trajectory (reviewed in³⁸ and elsewhere) with the possibility of binding to two receptors (ERα and ERβ) in some species. This itself generates modulation of estrogen’s function as the two can form a dimer. Binding of estrogen to its receptor(s) ultimately leads to trafficking to the nucleus where estrogen and its receptor can serve as transcription factors for several genes related to immune cell development and function. In addition, estrogen and its receptor can bind to the mitochondria and have effects related to not only gene expression but generation of ATP and decrease of ROS production. Finally, estrogen may bind to a membrane bound, G protein coupled receptor³⁹, which may mediate the estrogen signaling pathway by phosphorylating the estrogen bound ERα or ERβ and modify the capacity of the estrogen receptor/receptor complex to act as a transcription factor. Estrogen has also been shown to transcriptionally regulate gene expression not only in T and B cells, but also other immune cells such as dendritic cells⁴⁰, especially post activation. Expression of these receptors on natural killer cells⁴¹, and other cells such as eosinophils⁴² suggests broad-based regulation in the immune system. The resulting phenotype of estrogen responsiveness is likely complex. For example, estrogen may impair the negative selection of auto-reactive B cells, which could increase autoimmunity. In contrast, estrogen can also decrease B cell lymphopoiesis⁴³ and increase the expression of cytokines adaptive for the production of antibody responses. These responses can be highly protective of reproductive and gastrointestinal mucosal surfaces while inhibiting certain inflammatory responses which may be detrimental to developing fetal/placental tissues⁴⁴–⁴⁶.

Progesterone signaling in cells may be equally complex as several isoforms of the full-length isoform PR-B (e.g., A, C, M, T, S) have been described (reviewed in⁴⁷,⁴⁸). PR-A and PR-B are the main isoforms, which bind progesterone, traffic to the nucleus and control gene transcription (reviewed in⁴⁹ and elsewhere). The A and B forms may regulate each other’s action⁵⁰. The C isoform may bind progesterone, but does not contain a DNA binding domain, which suggests its function is to regulate the effect of circulating progesterone. The M form may be localized to the mitochondria and support progesterone-enhanced function⁵¹, such as

production of ATP. Low-affinity membrane-localized receptors for progesterone exist. These include a “family” of receptors,

(e.g., a, b, d, e, g) that may interact with molecules near the cell membrane which in turn activate signaling pathways and lead to molecular movement to the nucleus. Through these mechanisms, progesterone may indirectly mediate gene transcription at sites other than canonical progesterone response elements⁴⁹. Another class of progesterone binding membrane receptor with similar activity include those receptors with progesterone receptor membrane components (reviewed in⁴⁹). Furthering the complexity in progesterone receptor signaling pathways is the apparent cross talk between these and other steroid hormone signaling pathways⁵².

While nuclear progesterone receptors are present in thymic epithelium⁵³ and their expression modulates the environment in which T cells develop during pregnancy, the expression of such receptors in T cells, is controversial⁵⁴,⁵⁵. Evidence in different species however suggests that the regulatory effect of progesterone in T cells (e.g., driving generation of regulatory T cells or decreasing activation) may occur via non-canonical receptors⁵⁴,⁵⁶ or conversely via activation of glucocorticoid receptor pathways⁵⁷. By flow cytometry, a small proportion of circulating B cells express the “inhibitory” nuclear progesterone receptor A⁵⁸, perhaps as a method to downregulate inflammatory mechanisms in adverse pregnancies. However, evidence of membrane receptor expression in B cells is lacking⁵⁶. In contrast most B cells express the glucocorticoid receptor⁵⁹, raising the hypothesis that crosstalk between this and the progesterone receptor may be a mechanism for progesterone responsiveness in these cells.

Within the immune system, progesterone is thought to inhibit inflammatory immune responsiveness by diminishing pro-inflammatory cytokine production in peripheral blood leukocytes⁶⁰. The mechanism however may be via either direct or indirect transcriptional regulation of some cytokines, including IL-8, IL-6, and chemokines such as CXCL1, and CXCL2⁶¹. This may occur not only through effects on T and B cells, but also by decreasing the population of innate immune cells, such as neutrophils and NK cells, or by decreasing the function of macrophages and dendritic cells (reviewed in³⁷). Given the complexity of the timing and response to decreased progesterone and estrogen in lactating women, it is difficult to assert, based on these hormones, the likely inflammatory status of lactating versus non-lactating postpartum women.

BREASTFEEDING HORMONES AS IMMUNOMODULATORY MOLECULES

While the underlying mechanisms are incompletely understood, it is thought that hormones instrumentally involved in the breastfeeding process may regulate the immune system locally (in the breast) and systemically. The two most compelling hormones in this respect are prolactin (reviewed in⁶²) and oxytocin, which are in turn regulated by several other steroid hormones.

Prolactin is the hormone that induces synthesis of milk components like proteins and lipids. Prolactin receptors comprise many different membrane and soluble forms. The membrane-bound forms comprise homodimers of several different sizes (e.g., long, intermediate, short) depending on the size of the receptor extracellular or intracellular domain of the specific receptor (reviewed in⁶³). Short/long heterodimer forms can be inhibitory. Soluble forms lack transmembrane and intracellular domains. They chiefly mediate the availability and effective local concentration of prolactin but may also form dimers with other receptor forms and regulate their activity⁶⁴. Prolactin receptors are expressed on many cells throughout the immune system, including monocytes and macrophages⁶⁵, natural killer cells⁶⁶, and T cells⁶⁷. The expression of differing isoforms may be regulated by infectious or inflammatory stimuli. Adding to the complexity of signaling through these receptors is their low-level affinity for other members of this receptor class (reviewed in⁶⁸).

Prolactin’s binding to its receptor activates several different signal transduction pathways (e.g., JAK/STAT and MAPK), including those that induce gene expression and production of pro-inflammatory cytokines and chemokines in epithelial cells but also in macrophages⁶⁹. It also increases expression of the IL-2 receptor, through which it can affect the function of NK cells⁷⁰ and the maturation of CD4 and CD8+ thymocytes⁴⁵. In addition, prolactin can increase both viability and function in dendritic cells, presumably through regulation of NFKB⁷¹. Another of the hypothesized effects of prolactin on the immune system is dose-dependent bimodal (e.g., low versus high dose) regulation of the transcription factor T-bet⁷², a critical regulator of type 1 immune responses⁴⁵,⁷³. Additionally, a long form of the prolactin receptor has been identified in early bone marrow cells in animal models, suggesting a role in B cell development⁷⁴. It can rescue immature B cells from apoptosis, and inhibits B cell tolerance, while promoting B cell activation, proliferation and differentiation⁷⁴,⁷⁵. Through this it increases the generation of autoantibodies and autoimmune disease⁴⁵,⁷⁶–⁷⁹. Thus, prolactin is likely a broad-spectrum activator of postpartum immunity.

The neuropeptide oxytocin binds mainly to a G protein-coupled receptor that is found in several tissues of the central and peripheral nervous system. Interestingly, the spinal cord and brain – among other areas of the body – contain oxytocin neurons, which express receptors for immune cytokines including IL-6⁸⁰ and IL-1⁸¹ in response to cytokine binding with increased release of oxytocin. However, oxytocin itself drives regulation of immunity.

The oxytocin receptor can aggregate to form homodimers or heterocomplexes of two or higher order with other elements (e.g., the dopamine 2 receptor) and this likely regulates its downstream signaling effects (reviewed in⁸²). Oxytocin receptors are expressed on a wide variety of tissues and cells in the body, including those that participate in immunity⁸³. In dendritic cells, oxytocin signaling may promote a “tolerogenic” phenotype by decreasing some functions while increasing others through activation of the PI3K/AKT pathway⁸³. Similarly, macrophages express receptors for oxytocin⁸⁴. Signaling by oxytocin in these cells results in a relative inhibition of NF-kB and decreased expression of molecules such as IL-1b and TNFa while increasing tendency to express indicators of “M-2” polarization such as arginase⁸⁴. Inflammation increases expression of the oxytocin receptor, which then in the presence of the peptide decreases expression of inflammatory cytokines such as IL-6⁸⁵. T cells expressing the oxytocin receptor can respond to oxytocin by increasing intracellular calcium⁸⁶. This could have a wide-ranging effect on T cell activation (reviewed in⁸⁷) or effector function⁸⁸. However, enhanced premature activation within the thymus may support deletion of autoreactive T cells in favor of a mature response to environmental antigens. Though increased oxytocin expression by tumors increases B cell influx⁸⁹, evidence that oxytocin directly regulates B cell development or function is limited. That oxytocin release can be increased by inflammatory cytokines that it can in turn down modulate specific immune cells along, and that its levels fluctuate with breastfeeding suggest it may play a major role in postpartum immunity.

Animal models to examine the role of breastfeeding in the mother

Animal models have been used to delineate causality and specific mechanisms underlying disease processes as well as healthy physiology, and studies of the immune system relative to pregnancy are numerous⁹⁰–⁹². Studies specifically focused on breastfeeding are far fewer in number. Moreover, many animal studies of the immune system – including those which assess phenotype and activity of immune cells from peripheral blood relative to breastfeeding – compare animal mothers who are currently lactating to those who are not lactating, or to animals who have not been pregnant. Because pregnancy itself modifies the immune system and other elements of maternal physiology, this approach has made it somewhat difficult to specifically delineate the effects of breastfeeding. However, these studies are useful in the process of developing hypotheses to test in humans and in increasingly refined animal models. What follows is a discussion of some of the approaches to using animal models to address this question.

USE OF BROMOCRIPTINE

One strategy that has been employed to approach this question is the use of bromocriptine, a complex dopamine receptor agonist/antagonist, to prevent lactation. Administration of bromocriptine to dams during the early stages of pregnancy was used to observe treatment-associated postpartum care of their pups; the delineator used was a home cage versus a novel cage along with differential bromocriptine administration, but immune parameters were not investigated⁹³. The use of bromocriptine in animal models in studies concerning maternal immunity could enhance experimental design. Rather than using non-lactating, non-pregnant controls one could instead allow all animals to become pregnant and go through parturition and compare those treated with bromocriptine or vehicle control. Such an experiment might be a good parallel for bottle versus breastfeeding after pregnancy, though the use of this drug to inhibit milk production is confounded by its regulation of prolactin⁹⁴ (and reviewed in⁹⁵) which also may influence immunity.

COMPARISON OF EARLY VERSUS LATE TIME POINTS IN LACTATION

One technique used to assess how lactation may affect the immune system of lactating animals is collecting samples from various points throughout lactation, as well as before parturition or after lactation ceases⁹⁶. However, there appears to be no set standard for sample collection which allows for comparison across studies. One study done on dairy cows, for example, labeled animal subjects as being in lactation vs. dry periods to delineate subpopulations of T lymphocytes in mammary gland secretions⁹⁷, while another examined peripheral blood samples from postpartum and mid-late lactating animals to delineate CD8+ lymphocyte suppressor function⁹⁸. Another study euthanized pregnant mice on days 8 and 15 of pregnancy and day 8 postpartum to assess the difference in β1,4-Galactosyltransferase expression, an enzyme that increases the number of N-terminal galactose molecules on IgG molecules during pregnancy, potentially influencing their effector function⁹⁹. These techniques offer interesting insight into how the immune system of lactating animals changes postpartum, but do not specifically address the issue of breastfeeding/lactating versus non-breastfeeding/nonlactating mothers who have recently given birth.

REMOVAL OF PUPS/FORCED WEANING

While the literature includes studies in mice where the transfer of pups from one mother to another has been used to examine the effect on the pups’ immune system, little attention has been paid to the examination of either mother’s (biological or foster) immune system. A more informative approach might be to perform an experimental removal of the pups²³,³⁶. This would allow for the forced halting of milk production and prolactin release, albeit with a lag time of several days if not done immediately after parturition³⁶. This method would allow a comparison of lactation to non-lactation but may generate a level of stress that could be a confounder in its own right. Forced weaning and mammary gland involution in mice leads to transient increase in mammary monocytes, macrophages, dendritic cells and T cells, along with complex phenotypic and functional changes suggestive of immune tolerance within the dendritic and T cell pools. Exactly what changes occur with forced weaning in distal sites, such as the spleen are unclear³².

Removal of the nipples (thelectomy) could significantly inhibit lactation, (as does wearing a tight bra in humans¹⁰⁰) but is also confounded by changes in behavior and possible other effects in rats¹⁰¹ and mice¹⁰². An alternative in other species might be a temporary cessation of milking¹⁰³.

Observational studies in humans

Studies of specific phenotype and functional analysis of peripheral blood or local or systemic lymphoid tissue from healthy mothers who exclusively breastfed

their infants, when compared with mothers who exclusively formula-fed are relatively rare. Limited evidence suggests that exclusive breastfeeding is associated with a lower peripheral blood CD3+ cell percentage, but higher serum γ-IFN, higher peripheral blood cell response to in vitro stimulation with PHA, and lower IL-10 production by PBMC in response to stimulation at 4–6 weeks postpartum, and lower symptoms of infection, suggesting that breastfeeding may play a positive role in maternal immunity¹⁵. However, some evidence suggests that in states of chronic infection, it is the postpartum state and not breastfeeding per se, that alters immune and inflammatory parameters¹⁰⁴. Although studies directly observing breast versus bottle-feeding women is limited, some information may be gained by examining certain pathologic states.

BREASTFEEDING AND RISK OF BREAST CANCER

Breast cancer is the most diagnosed cancer in women in the U.S. and one of the most significant contributors to cancer deaths worldwide¹⁰⁵–¹⁰⁷. Immune system involvement may occur early in tumorigenesis, leading to removal of cancer cells, or late, perhaps decreasing metastasis. The role of the immune system in modifying breastfeeding and breast cancer is likely complex and confounded by diet, hormonal factors, toxin exposure, and other variables such as body weight and vitamin D status¹⁰⁷–¹⁰⁹. Studies suggest that breastfeeding may reduce the risk of breast cancer¹⁰⁶.

Particularly aggressive forms of breast cancer occur in the peripartum period, including cases that are triple-negative for estrogen, progesterone, and HER2 receptors¹⁰⁵,¹⁰⁷. These tumors have a high capacity for metastasis, since they may occupy or utilize the neo-lymphangiogenesis that occurs as the breast is involuting (reviewed in¹¹⁰). It is said that the involuting breast environment supports inflammation¹¹¹,¹¹² and tumorigenesis¹¹³. This environment may depend on mast cell presence¹¹⁴, and B cell presence, phenotype and function³³, and on infiltrating macrophage phenotypes, which have an immune “suppressive” phenotype³¹, and which drive neo lymphangiogenesis¹¹⁰. T cells specific for breast cancer neoantigens have been described, and moreover, pregnancy in a mouse model has been shown to generate antigen specific T cells that in turn decrease disease¹¹⁵. Consistent with this is the relative lack of both CD4 and CD8 T cells in the involuting breast tumor microenvironment³¹, and the high ratio of macrophages to T cells. These data may support the thought that the immune system is a link between pregnancy and breast cancer and further supports examination of the role of the immune system and breastfeeding in breast cancer risk.

ALLERGIC DISORDERS

While a large body of literature has investigated the impact of maternal allergy, diet and breastfeeding on subsequent allergic disease in the offspring, little data exists on the effect of breastfeeding itself on maternal allergic disease. While breastfeeding can reveal or precipitate rare extremes of allergy¹¹⁶, it is currently unclear if breastfeeding changes the course of chronic allergic disease in the mother¹¹⁷.

AUTOIMMUNE DISORDERS

Autoimmune disorders result from a disruption, either acute or chronic, of tissue homeostasis that leads to an immune response. These disorders primarily affect women of reproductive age. Evidence from several studies is that autoimmune disorders relapse or abate during pregnancy and in the postpartum period recur, flare, or become clinically evident in previously non-symptomatic women¹¹⁸–¹²⁵. This combined with the observations combined with the sexually dimorphic nature of the disorders¹²⁶ suggests regulation of the disorders by the sex chromosomes and their gene products.

One particularly devastating but rare pregnancy-related autoimmune disease is peripartum cardiomyopathy¹²⁷. Though a mediator of this disease is thought to be a form of prolactin, there has been interest in the role of breastfeeding modification of the disease due to potential effects on the immune system. Interestingly, data suggests that in women with the disease, breastfeeding increased prolactin, as expected, but in addition increased circulating CD8 T cells¹²⁸. Circulating CD4 T cells were lower with breastfeeding. Breastfeeding did not significantly affect disease outcome as cardiac function was only slightly higher in breastfeeding versus non-breastfeeding patients¹²⁸. Most data on breastfeeding and auto-immunity, however, comes from observation of only a handful of well-known autoimmune diseases.

Inflammatory Bowel disease is an example of a disease with increased risk for postpartum flares¹²⁷, although this may be related to therapy de-escalation¹²⁹. Little data on the role played by breastfeeding in postpartum exists but suggests there may be a protective effect in some diagnoses¹³⁰.

Systemic lupus erythematosus (SLE) is a chronic autoimmune disease with a variety of symptoms that has been shown to worsen during both pregnancy and the postpartum period particularly in patients with active disease prior to pregnancy (reviewed in¹³¹). There is particular interest in the field on how prolactin, which contributes to SLE pathogenesis and activity, may contribute to the worsening of disease activity¹³¹. Conflicting data exists. For example, one study of patients early postpartum (6 weeks) suggested that formula alone feeding was associated with higher disease activity as compared to breastfeeding (alone or with formula supplementation)¹³², regardless of immune suppressive therapy. However, another study found no significant association¹³³, though there was observed a trend to increased disease severity in non-breastfeeding mothers at one-year post-partum. Small numbers and important confounders (e.g., smoking¹³⁴ and the effect of time¹³⁴ of breastfeeding) may blur the association. There moreover may be inherent selection bias in such data, as mothers may be hesitant to breastfeed due to fear of the safety profile of drugs used to treat the disease¹³⁵. Larger studies of postpartum patients exist¹³¹, but do not specifically address breastfeeding. While it is possible to assert that lack of breastfeeding may lead to increased disease activity, more research is needed in this area, including refinement of specific disease activity metrics delineated and perhaps a meta-analysis of existing data.

Rheumatoid arthritis (RA) is a complex disease in that many patients experience a decrease in disease severity while pregnant¹¹⁹, while RA flares in¹²⁰ or may present with new onset disease¹³⁷ in the postpartum. The relationships between breastfeeding and hormones are complex, e.g., prolactin and RA are complex, increasing the difficulty in understanding the association, if any, with breastfeeding. A 2015 meta-analysis found an inverse relationship between ever breastfeeding and subsequent RA development, with decreased subsequent risk in both women who breast fed for shorter (<12 months) or longer (>12 months) time frames¹³⁸, however this may be related to several factors not related to immune regulation. Studies specifically measuring the effect of breastfeeding on postpartum disease severity are rare, as many studies have focused on fear of medication and intent to breastfeed, and other sociodemographic factors supportive of breastfeeding (for example¹³⁹). New tools to measure disease severity in the context of pregnancy may be better able to provide an assessment¹⁴⁰ of disease, and specific information on immune parameters.

Multiple sclerosis (MS) is another autoimmune disorder that more frequently affects women. Evidence suggests that while disease severity decreases during pregnancy, patients experience flares or worsening disease in the postpartum period¹¹⁹, with 30% of women with MS experiencing a relapse in the first 3 months postpartum¹²². The relationship between breastfeeding hormones (e.g., prolactin) and RA are complex, increasing. Many relevant disease-modifying therapies are not recommended in pregnancy or lactation, so many women choose to return to these therapies and forgo breastfeeding in order to prevent relapses, particularly if pre-pregnancy disease was severe. Conversely, those with mild disease could choose to not restart medication and go on to breastfeed. An early large prospective study of pregnant women with MS suggested that breastfeeding did not have a protective effect against postpartum disease severity (flares)¹²¹. Nearly two decades and several studies later, a meta-analysis found a protective effect. Subsequently, a meta-analysis found that women who do choose to breastfeed have at least a 37% less chance of postpartum relapse compared with women who don’t¹²². The benefit of breastfeeding was found to be stronger in the analysis of studies which required at least two months of exclusive breastfeeding than in the analysis including studies which allowed nonexclusive breastfeeding¹²². This may be because studies in the analysis differed with respect to the effect of nonexclusive breastfeeding with one study reporting that women who breastfed nonexclusively had comparable relapse risk to women who did not breastfeed at all. A second meta-analysis that same year found that population rates of MS relapse were not related to the proportion of women in the population who breastfed¹⁴², and this finding may have been related to decreasing rates overall. Finally, in the era of more widespread use of pre-conceptional disease modifiers, a recent meta-analysis found that the use of such drugs pre-conceptionally was associated with increased and exclusive breastfeeding and also associated with decreased postpartum risk¹⁴³. None of these analyses specifically looked at systemic immune parameters or metrics of systemic immunity (e.g., response to vaccines). Once again, more research is needed concerning progression and flares of this autoimmune disease and breastfeeding.

Though studies centering on these diseases could give us significant insight into the relationship between breastfeeding and the immune system, they are confounded by disease state and treatment using and are limited by the detail of the functionality and phenotype of the immune cells observed.

Conclusion

Breastfeeding is a critical driver of postpartum physiology in animal models¹⁴⁴ and in humans⁶,⁸,¹⁰–¹³. Breastfeeding is associated with hormones that may have a regulatory effect on immunity. The lack of breastfeeding may lead to a state of postpartum residual, low-level estrogen, which may be inflammatory, in the presence of falling or extremely low-level progesterone and lack of the immunoregulatory activity of breastfeeding hormones such as prolactin and oxytocin. It is hypothesized that this milieu may change immune cell development, homeostasis and effector function, but it may also lead to increased tissue (especially the breast) dysregulation which could in turn fuel inflammatory processes.

Studies examining the association between breastfeeding and postpartum flares of autoimmune disease are very complicated and confounded by such issues as baseline disease status, the availability of immunosuppressive drugs and biologics, and the changing overall presence of disease. They further comprise very little data on specific immune parameters. Further study of the development of breast cancer in young postpartum women is likely to be informative, but there still needs to be effort in the detailed examination of breast versus bottle feeding women after a normal pregnancy, as these may be less confounded by exogenous exposures, medications, and other interventions.

Finally, there is still a need for research using well-controlled experiments in animal models to delineate lactation-associated phenotypic and functional changes in immune cell populations as well as underlying mechanisms.