Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Feb 1.
Published in final edited form as: Am J Phys Anthropol. 2018 Dec 21;168(2):329–339. doi: 10.1002/ajpa.23752

Buffered or impaired: Maternal anemia, inflammation and breastmilk macronutrients in northern Kenya

Masako Fujita 1,2, Nerli Paredes Ruvalcaba 1,2, Katherine Wander 3,4, Mary Corbitt 1,2, Eleanor Brindle 5
PMCID: PMC6352968  NIHMSID: NIHMS998840  PMID: 30575959

Abstract

Background:

Maternal anemia has adverse consequences for the mother-infant dyad. To evaluate whether and how milk nutrient content may change in ways that could “buffer” infants against the conditions underlying maternal anemia, this study assessed associations between milk macronutrients and maternal iron-deficiency anemia (IDA), non-iron-deficiency anemia (NIDA), and inflammation.

Methods:

A secondary analysis of cross-sectional data and milk from northern Kenya was conducted (n=204). The combination of hemoglobin and transferrin receptor defined IDA/NIDA. Elevated serum C-reactive protein defined acute inflammation. The effects of IDA, NIDA, and inflammation on milk macronutrients were evaluated in regression models.

Results:

IDA (β=0.077, p=0.022) and NIDA (β=0.083, p=0.100) predicted higher total protein (ln). IDA (β=−0.293, p=0.002), NIDA (β=−0.313, p=0.047), and inflammation (β=−0.269, p=0.007) each predicted lower fat (ln); however, anemia accompanying inflammation predicted higher fat (β=0.655, p=0.007 for IDA and β=0.468, p=0.092 for NIDA). NIDA predicted higher lactose (β=1.020, p=0.003).

Conclusions:

Milk macronutrient content both increases and decreases in the presence of maternal anemia and inflammation, suggesting a more complicated and dynamic change than simple impairment of nutrient delivery during maternal stress. Maternal fat delivery to milk may be impaired under anemia. Mothers may buffer infant nutrition against adverse conditions or poor maternal health by elevating milk protein (mothers with IDA/NIDA), lactose (mothers with NIDA), or fat (mothers with anemia and inflammation). This study demonstrates the foundational importance of maternal micronutrient health and inflammation or infection for advancing the ecological understanding of human milk nutrient variation.

Keywords: iron deficiency anemia, non-iron deficiency anemia, infection, transferrin receptor, C-reactive protein

Introduction

Anemia is a global public health concern, affecting numerous countries regardless of income level (Miller, 2013). Anemia is defined as a low blood concentration of hemoglobin, the oxygen-transport protein in the red blood cell. Consequences of anemia include diminished work capacity and impaired cognitive function (Christian, 2005; Murray-Kolb & Beard, 2007). Maternal anemia is further associated with increased risk of low birth weight, and elevated maternal and perinatal mortality, and in severe cases of anemia, with child mortality (Christian, 2005; Murray-Kolb & Beard, 2007; Picciano, 2003). Recent studies report altered quality of breast milk in association with maternal anemia, including reduced immunological components (França et al., 2013) and minerals (El-Farrash, Ismail, & Nada, 2012), and altered nutritional contents (Corbitt, Paredes Ruvalcaba, & Fujita, 2018a, 2018b; França et al., 2013; Fujita, Brindle, & Lo, 2015). The altered quality of breast milk may be an important pathway linking maternal anemia to child malnutrition, morbidity and mortality in some environments. The effect of maternal anemia on milk nutrients, however, is still poorly understood, particularly for macronutrients; studies report increases, decreases, or no change in milk macronutrient contents in association with maternal anemia (El Farrash et al., 2012; França et al., 2013; Corbitt et al., 2018a, 2018b).

The variable effect of maternal anemia on milk macronutrients may be better understood by applying the perspective of maternal buffering, or more specifically in terms of the varying extent to which breast milk nutrient content may be buffered from the effect of maternal malnutrition or other health issues underlying anemia. It may be that seemingly counter-intuitive lack of associations between maternal anemia and milk content, or even higher nutrient levels reported by some studies, reflect milk being buffered from the effect of relatively mild maternal malnutrition. This buffering may not be feasible for mothers with severe undernutrition or other serious health issues underlying anemia who would not have biological means. In such cases, impairments in milk nutrients may occur in association with maternal anemia.

The maternal buffering hypothesis (Pond, 1977; a thorough review in Quinn, Diki Bista, & Childs, 2016), suggests that nutrients in mammalian milk may be buffered against impairments due to moderate environmental stress and maternal nutrition primarily mediated by maternal body fat store. We apply this hypothesis to the question of maternal anemia: instead of maternal energy balance per se, we utilize a broad concept of maternal functional nutrition/health – consisting of the complex interplay of macronutrient stores, micronutrient stores, and other health conditions (e.g. presence of infections) that can cause anemia when jeopardized – as the foundational resource that enables mothers to buffer milk nutrients. The objective of the study therefore was to evaluate whether and how milk nutrient content may differ in ways that could “buffer” infants against the conditions underlying maternal anemia. We assessed associations of milk macronutrients with maternal iron-deficiency anemia (IDA), non-iron-deficiency anemia (NIDA), and inflammation. Distinguishing the types of anemia may offer insights to the inconsistent associations between maternal anemia and milk macronutrients.

Background

Maternal anemia, malnutrition and milk macronutrients: varied associations

The existing literature seems to agree that anemia tends to compromise milk micronutrients (El-Farrash et al., 2012 for minerals) and immunological factors (França et al., 2013 for antibodies and complement proteins) while the effects on macronutrients are more variable. In Brazil, milk protein was consistently elevated in anemia regardless of the type of milk (colostrum, transitional, and mature milk), while milk fat was elevated in the colostrum but reduced (non-significantly) in transitional or mature milk (França et al., 2013). Our previous studies among Kenyan women similarly found that maternal anemia was associated with higher milk protein and lower milk fat (Corbitt et al., 2018a, 2018b; all mature milk) while maternal inflammation may alter this association (Fujita et al., 2015).

The varied associations between milk macronutrients and maternal anemia parallel the wider literature on maternal undernutrition: Lönnerdal’s seminal review published in 1986 argues that maternal malnutrition generally has little effect on total protein concentrations of milk. More recent studies, however, show a positive (Chang et al., 2015; Grote et al., 2015; Michaelsen, Skafte, Badsberg, & Jørgensen, 1990) or a negative (Bachour, Yafawi, Jaber, Choueiri, & Abdel-Razzak, 2012) association between maternal protein-energy status (BMI) and milk protein. Milk fat is similarly reported to have a positive association (Chang et al., 2015; Grote et al., 2015; Michaelsen et al., 1990; Nommsen, Lovelady, Heinig, Lönnerdal, & Dewey, 1991) or no association (Bachour et al., 2012; Mandel, Lubetzky, Dollberg, Barak, & Mimouni, 2005) with maternal protein-energy deficiency. Milk lactose has been described as the least affected by maternal nutritional status, yet a recent study reports a significant inverse relationship between BMI and lactose (Chang et al., 2015). Severe undernutrition may be associated with a decrease in milk volume (Jelliffe & E. Jelliffe, 1978) and multiple nutrients (Prentice et al., 1994).

Complexity of malnutrition

These varied associations between maternal malnutrition and milk macronutrient contents may be attributable to the complexity of malnutrition (Lönnerdal, 1986): Malnourished individuals are often simultaneously deficient in multiple nutrients (e.g., poor diets may be deficient in protein, energy and micronutrients). Malnourished individuals may have sufficient macronutrient intake but may be deficient in specific micronutrients (e.g., minerals, vitamins) that may interrupt normal nutrient metabolism (e.g., insufficient vitamin B6 hinders protein metabolism; Lönnerdal, 1986). Furthermore, malnourished individuals are often more susceptible to infectious diseases. The effect of malnutrition may be confounded by the effect of infectious disease processes, or there may be malnutrition-infection interaction. The nutrient-nutrient or nutrient-infection interactions may also contribute to some of the inconsistency in the literature on milk macronutrients and maternal malnutrition, including anemia.

Maternal buffering hypothesis reconceived: functional nutrition and health

The varied results for milk macronutrients may also be attributable to maternal buffering, or more specifically to the varying extent to which breast milk content may be buffered from the effect of maternal malnutrition. The maternal buffering hypothesis, in its original formulation, predicts that macronutrients in human milk remain constant relative to fluctuations in maternal nutrition (Quinn, Largado, Power, & Kuzawa, 2012; Quinn, Diki Bista, Childs, 2016) with exception for milk fat, which may increase with maternal adiposity (Prentice, Goldberg, & A. Prentice, 1994). This hypothesis is based on the idea that maternal fat stores will allow mothers to buffer milk nutrients (and thus protect infants against undernutrition) in harsh environments (Pond, 1977) because adipose tissue can serve as a secondary source of milk energy when dietary energy is in shortage (Quinn, Diki Bista, Childs, 2016). Given the complexity of malnutrition, however, we expect that the extent of milk buffering may vary from complete, to incomplete, to none at all, depending on the severity/type of maternal malnutrition (e.g. micronutrient deficiencies), the state of health (infection, inflammation, etc.), and/or the physiological ease of delivering a nutrient to milk. Here, the maternal buffering hypothesis is utilized grounded in the notion of maternal functional nutrition and health, with an emphasis on micronutrient health and immunological state instead of relying solely on maternal energy. This builds on the important observation that some micronutrients, such as iron, are of crucial importance on a par with energy as the foundational resource for human reproduction (Miller, 2016), and that our knowledge can be advanced by being attentive to micronutrient health (Fujita, 2008; Fujita et al., 2011; Hinde et al., 2013; Miller, 2010).

If delivery of some nutrients to milk can more easily be increased than others under nutritional stress, then we may expect an increase in one nutrient to compensate for the decrease in another nutrient, particularly if mothers have relatively mild malnutrition. The ease at which the maternal body can adjust the milk nutrient level may differ by nutrient origin. Many nutrients in human milk originate primarily in maternal blood while others are primarily synthesized in the mammary. Milk fat and lactose belong to the former, primarily originating in maternal blood lipids and glucose, respectively (Aumeistere, Ciporviča, Zavadska, & Ceļmalniece, 2017; Murase et al., 2009). Milk protein arises primarily from mammary synthesis (Murase et al., 2009). Increasing the rate of protein synthesis might be relatively easier than increasing uptake of lipids or glucose. If so, we may see no change (buffered) or increased protein levels associated with maternal anemia while milk fat and lactose may be compromised. In contrast, mothers suffering severe malnutrition or specific nutritional deficiencies (e.g., iron deficiency) may not be able to buffer milk. In such situations, we may expect to see malnutrition associated with lower levels of all macronutrients. By distinguishing the type or severity of malnutrition, it may be possible to bring some clarity to the effect of maternal “malnutrition” on milk macronutrients and contribute to clarification of the principles of maternal buffering.

No research to date has investigated potential changes in breast milk macronutrients by distinguishing different types of anemia. This may be because it can be difficult to distinguish different types of anemia in non-clinical settings (Darnton-Hill, Paragas, & Cavalli-Sforza, 2007). Distinguishing different types of anemia may offer new insights to the inconsistent associations between maternal anemia and milk macronutrients. Anemia can arise via multiple pathways, including blood loss occurring during childbirth; loss or destruction of red blood cells due to infectious diseases; micronutrient malnutrition (e.g., iron, folate, vitamin B12, and/or vitamin A deficiency); or, chronic activation of iron withholding mechanisms (see below). The relative importance of these causes of anemia varies widely across nutritional and infectious disease ecologies.

Types of Anemia

Iron deficiency anemia (IDA) is the most common type of nutritional anemia (WHO, 2008). IDA develops when there is inadequate iron intake or iron store to support normal production of red blood cells (Miller, 2013). Elevated iron requirements during pregnancy can be a major contributor to IDA among postpartum mothers; IDA is common among pregnant and postpartum women as a result of high iron requirements to facilitate prenatal transfer of iron to the fetus (Miller, 2016). IDA can also arise from chronic or repeated activation of the iron-withholding response. Iron withholding is part of the acute phase response (APR) to infectious disease: iron absorption in the gut is reduced and iron sequestration is increased by multiple mechanisms, including iron binding to ferritin and turnover of senescent red blood cells in the spleen (Jurado, 1997; Kent, Weinberg, & Stuart-Macadam, 1994; Weinberg, 1984). These mechanisms combat infection by reducing the availability of iron to infectious agents (Cassat & Skaar, 2013; Weinberg, 1984). When their activation is prolonged (e.g., in the case of chronic infectious disease or an inflammatory disorder; Jurado 1997), however, they can cause anemia in individuals whose iron intake would otherwise be adequate to meet their needs (Weinberg, 1984).

Non-iron deficiency anemia (NIDA) includes anemia due to blood loss or red blood cell destruction, for example, resulting from infectious disease processes (particularly hookworm; Darnton-Hill et al., 2007; Lynch, 2007; Thurnham & Northrop-Clewes, 2007; Thurnham, 2014), as well as anemia resulting from micronutrient deficiencies other than iron deficiency. NIDA may also include cases with rarer chronic diseases such as autoimmune hemolytic anemia or bone marrow cancer; however, it is unlikely that these are important in explaining anemia in the current study in northern Kenya among seemingly healthy mothers.

In environments with high disease load and food insecurity, multiple types of anemia may co-exist. Anemic individuals may be heterogeneous in terms of the primary causation of anemia, with some suffering anemia due to nutritional deficiencies or due to infection, while others have anemia from the combined effects of nutritional deficiencies, infection, and inflammation; similarly, infectious disease may be contributing to anemia both directly and through inflammatory pathways restricting the uptake, availability, and recycling of iron (Drakesmith & Prentice, 2012; Thurnham, 2014).

Maternal inflammation or infection without anemia has been associated with premature birth and low birth weight (Rogers & Velten, 2011). The research on the associations between maternal inflammation/infection and milk components is inconsistent. Some studies suggest that maternal inflammation/infection may alter milk fatty acids or immune components (Gardner et al., 2017; Groer, Davis, & Steele, 2004; Mizuno et al., 2012), while others found no associations of subclinical inflammation with milk micronutrients (Fujita, Lo, & Brindle, 2017; Zavaleta et al., 1995).

It is unclear if these different types of anemia, with or without inflammation, may equally impact breast milk macronutrient contents. Studying different types of anemia may shed light on potential pathways linking maternal anemia to breast milk nutrients. The present study evaluated the associations of maternal anemia with levels of macronutrients in breast milk, distinguishing IDA, NIDA, and allowing for interactions between anemia and elevated inflammation. Table 1 summarizes the definitions of IDA/NIDA utilized in this study and possible causes for each anemia type.

Table 1.

Definition of IDA/NIDA and possible causes

Anemia type Definition Possible causes
IDA Hb<12 g/dl; DBS TfR>5 mg/l •Inadequate iron intake or iron store
•High iron usage for pregnancy
•Chronic/repeated activation of the iron-withholding response to infectious disease/inflammation
NIDA Hb<12 g/dl; DBS TfR≤5 mg/l •Blood loss due to childbirth, injury, infectious disease
•Red blood cell destruction by infectious disease processes (e.g., hookworm)
•Micronutrient deficiencies other than iron (e.g., vit. A, B9, B12)
Open in a new tab

IDA Iron deficiency anemia; NIDA Non-iron deficiency anemia

Hb: hemoglobin; DBS TfR: dried-blood spot transferrin receptor

Predictions

“Impairment” should be apparent as lower milk macronutrient among mothers with anemia (or inflammation). “Buffering” should be apparent as no change in milk macronutrient among mothers with anemia (or inflammation).

Impairment and buffering differ by macronutrient:

The extent that mothers can buffer milk may differ by nutrient, because some milk nutrients may be more easily increased than others under stress. Similarly, buffering may involve compensation for decreases in one macronutrient with increases in another, in which case we predict higher concentrations of some macronutrients associated with anemia (or inflammation). Because milk fat and lactose primarily originate from maternal blood lipids and glucose, respectively (Aumeistere, Ciporviča, Zavadska, & Ceļmalniece, 2017; Murase et al., 2009), while milk protein arises primarily from mammary synthesis (Murase et al., 2009), we predict milk fat and lactose delivery to milk are more vulnerable to impairment due to anemia (or inflammation), while milk protein is more likely to be buffered or increased due to maternal anemia (or inflammation).

Impairment and buffering differ by anemia types and inflammation:

Iron deficiency and non-iron deficiency anemia reflect different underlying disease states, which may differently affect delivery of macronutrients to milk and maternal capacity for buffering. More severe malnutrition or iron deficiency may limit a mother’s capacity to buffer milk. Elevated inflammation may be a source of additional energetic demands for mothers (fighting infection) and may contribute to iron deficiency. Thus, we predict mothers with IDA, those with inflammation, and those with anemia in combination with inflammation are more vulnerable to impairment.

Methods

Cross-sectional data and cryogenically archived milk specimens, originally collected in a 2006 study with Ariaal agro-pastoralists of northern Kenya, were utilized (n=204). The internal review boards of the University of Washington and Kenya Medical Research Institute approved the original data/specimen collection (Fujita, 2008). All participants provided informed consent. The ecological and sociocultural contexts of Ariaal agro-pastoralists have been described in depth elsewhere (Fratkin, 2004; Fratkin & Roth, 1990, 2005; Fratkin & Smith, 1995; Fujita, 2003; Fujita, Roth, Nathan, & Fratkin, 2004, 2005; Shell-Duncan & Yung, 2004; Roth & Fratkin, 2005; Roth & Ngugi, 2005). At the time of data collection, northern Kenya suffered a serious drought. The loss of agricultural crops and livestock due to drought created a high level of food insecurity and hunger for people in the region. Many of the study participants were aided by drought relief distributions (Fujita, 2008).

The mothers were a convenience subsample of a stratified random sample of 241 mothers originally selected for a study on breast milk vitamin A (Fujita, 2008; Fujita et al., 2011). The original study surveyed seemingly healthy breastfeeding mothers, excluding those who exhibited clinical symptoms of acute infection such as fever and vomiting (for full description of original sampling method and exclusion criteria, see Fujita et al., 2011). For the present study, the additional exclusion criteria were being <18 years of age and the lack of complete information for the following variables: milk macronutrients, age, parity, time postpartum, resumption of menses, breastfeeding frequency, BMI, hemoglobin, C-reactive protein (CRP), and transferrin receptor (TfR). The remaining 204 mothers and the original 241 mothers did not differ significantly with respect to their age, BMI, time postpartum, or hemoglobin.

Variables

Milk nutrients

Milk fat concentration data were available from our previous research (Fujita, 2008). The data were based on creamatocrit (Fujita, 2008; Lucas, Gibbs, Lyster, & Baum, 1978; Wang, Chu, Mellen, Shenai, 1999) of thawed milk specimens (foremilk collected after an overnight fast), converted into lipid concentrations (g/dl) using the formula: Lipid = 0.54creamatocrit + 0.39. Lactose concentrations were determined using an enzyme-based colorimetric lactose assay kit (BioAssay Systems, Hayward, CA, Cat. No. ELAC-100), following the manufacture recommended protocol. Total protein was determined using a Micro BCA Protein Assay Kit (Thermo, Cat.23235), following the kit protocol except for the lower incubation temperature (25°C instead of 36°C) to prevent over-reaction. Both lactose and total protein were determined in triple-centrifuged milk sera in the Biomarker Laboratory for Anthropological Research at Michigan State University. The intra-assay CVs were <10% and the inter-assay CVs were <15% for control specimens of low and high concentrations across 14 plates for lactose and 7 plates for total protein.

Maternal anemia, iron status, and inflammation/infection

Low hemoglobin levels (<12 g/dl) defined anemia (Nestel, 2002). Elevated DBS TfR (>5 mg/l) defined iron deficiency (Wander, Shell-Duncan, & McDade, 2009; Fujita & Wander, 2017). IDA was defined as anemia in presence of iron deficiency. NIDA was defined as anemia in absence of iron deficiency. Elevated inflammation (probably indicative of acute infection) was identified with elevated blood serum C-reactive protein (>5 mg/l; Brindle, Fujita, Shofer, & O’Connor, 2010; Freedman, 2001; Fujita & Wander, 2017; Nakagomi, Freedman, & Geczy, 2000).

Other variables

Factors reported to have associations with milk macronutrients include: BMI, complementary feeding, age, parity, menses, time postpartum, and milk volume (Grote et al., 2015; Dewey et al., 1984; Nommsen et al., 1991; Mandel et al., 2005; Emmett, 1997; Michaelsen et al, 1990). The same for anemia include BMI, age, parity, breastfeeding status, and inflammation/infectious disease (Alene & Dohe, 2014; Antelman et al., 2000;Bodnar, Scanlon, Freedman, Siega-Riz & Cogswell, 2001; Semba et al., 2002). These variables except milk volume were available from our previous research (Fujita, 2008; Fujita et al., 2011, 2015, 2017; Fujita, Brindle, Lo, Castro, Cameroamortegui, 2014; Fujita & Wander, 2017; Stone & Fujita, 2016) and utilized as covariates to adjust for their effects on milk characteristics.

Statistical Tests

Regression models for milk characteristics were constructed with IDA, NIDA, and inflammation as predictors, along with IDA/NIDA-inflammation interaction terms. The models had the following form: Milk nutrient = β0 + β1 IDA + β2 NIDA + β3 Inflammation + β4 IDA × Inflammation + β5 NIDA × Inflammation + covariates For the outcome, natural log values were used for total protein and milk fat to remedy their skewed distributions. Control covariates were BMI, age, time postpartum, parity, resumed menses, and breastfeeding frequency. These covariates were evaluated for interaction with IDA and NIDA, without other interaction terms, and rejected if the probability for the interaction was not significant. Stata V.13 was utilized for statistical computation and graphic summary of results. The α-level was set at .05.

Results

Sample characteristics are shown in Table 2. The mean (± SD) age was 28 ± 7 years ranging from 18 to 46 years, BMI was 19.7 ± 3 ranging from 14.4 to 33.5, and postpartum time was 244 ± 135 days spanning from 24 to 585 days. The percentage of mothers with IDA was 17.6% (n=36), 6.9% of mothers had NIDA (n=14), and 17.2% had elevated inflammation (n=35).

Table 2.

Sample characteristics

Overall (n=204) IDA (n=36; 17.6%) NIDA (n=14; 6.9%) Inflammation (n=35; 17.2%)
Maternal characteristics
Age (year) 28 ± 7 26 ± 5* 27 ± 8 27 ± 6
BMI 19.7 ± 2.9 19.6 ± 2.1 20.2 ± 2.9 19.9 ± 3.9
Time postpartum (day) 244 ± 135 208 ± 136† 221 ± 137 252 ± 132
Parity 3.6 ± 2.2 2.80± 1.9* 2.9 ± 1.7 3.4 ± 2.2
Breastfeeding freq./day 9.2 ± 4.1 9.3± 4.0 9.8 ± 4.5 10.3 ± 5.1
Resumption of menses 11% 8% 7% 6%
Inflammation 17% 14% 36% 100%
Iron deficiency 31% 100% 0% 23%
Milk nutrients
Total protein (g/dl) 0.99 ± 0.2 1.07 ± 0.2* 1.06 ± 0.2 0.95 ± 0.2
Total protein (ln) −0.03 ± 0.2 0.05 ± 0.2** 0.04 ± 0.2 −0.07 ± 0.2
Fat (g/dl) 2.6 ± 1.3 2.3 ± 1.4 2.1 ± 0.6 2.3 ± 1.0
Fat (ln) 0.8 ± 0.5 0.7 ± 0.5* 0.7 ± 0.3 0.8 ± 0.4
Lactose (g/dl) 7.8 ± 1.2 7.7 ± 1.2 8.8 ± 1.0** 7.9 ± 1.1
Open in a new tab

IDA Iron deficiency anemia; NIDA Non-iron deficiency anemia

†

p<.10,

*

p<.05,

**

p<.01 for two-sided t-test compared to their counterparts

In regression models, milk fat was significantly lower among anemic mothers (Table 3 fat interaction model; Figure 1). The magnitude of this effect was similar for mothers with IDA (β1=−0.293) and NIDA (β2=−0.313), and was significant for both categories of anemia (p=0.002 and p=0.047, respectively). Milk fat was also significantly lower among mothers with elevated inflammation (β3=−0.269, p=0.007). There was significant interaction between anemia and elevated inflammation, which essentially reversed the effect of anemia on milk fat—among mothers with elevated inflammation, anemia was associated with higher milk fat; this interaction was significant for IDA (p=0.007) and marginal for NIDA (p=0.092).

Table 3.

Regression models for macronutrients in breast milk with and without interaction

Interaction model Main-effect model
β SE P β SE P
Total Protein (ln)
IDA 0.074 0.036 0.042 0.077 0.033 0.022
NIDA 0.032 0.061 0.597 0.083 0.050 0.100
Inflammation −0.062 0.039 0.110 −0.040 0.033 0.226
IDA*Inflammation 0.033 0.094 0.722 — —
NIDA*Inflammation 0.154 0.108 0.155 — —
Fat (ln)
IDA −0.293 0.092 0.002 −0.202 0.087 0.022
NIDA −0.313 0.157 0.047 −0.178 0.130 0.174
Inflammation −0.269 0.099 0.007 −0.120 0.087 0.167
IDA*Inflammation 0.655 0.240 0.007 — —
NIDA*Inflammation 0.468 0.276 0.092 — —
Lactose
IDA −0.088 0.243 0.717 −0.078 0.225 0.729
NIDA 1.021 0.413 0.014 1.020 0.335 0.003
Inflammation −0.033 0.260 0.898 −0.022 0.223 0.922
IDA*Inflammation 0.074 0.633 0.907 — —
NIDA*Inflammation 0.003 0.728 0.996 — —
Open in a new tab

IDA Iron deficiency anemia; NIDA Non-iron deficiency anemia

Models are adjusted for covariates: BMI, age, time postpartum, parity, resumed menses, and breastfeeding frequency. Time postpartum was a significant (p<.05) positive predictor for fat and a significant negative predictor for lactose and protein. Parity was a significant negative predictor for fat, and breastfeeding frequency was a significant negative predictor for protein and a marginal positive predictor for lactose.

graphic file with name nihms-998840-f0001.jpg

Open in a new tab

Milk lactose was significantly higher for mothers with NIDA (β2=1.020; p=0.003; compared to non-anemic mothers), but not for mothers with IDA (Table 3 lactose main-effect model; Figure 2). Milk lactose was not different for mothers with and without elevated inflammation; there was no significant interaction between anemia and elevated inflammation as predictors of milk lactose.

graphic file with name nihms-998840-f0002.jpg

Open in a new tab

Milk protein was higher for anemic mothers than non-anemic mothers (Table 3 protein main-effect model; Figure 2). The magnitude of this effect was similar for mothers with IDA (β1=0.077) and NIDA (β2=0.083); the effect was significant (p=0.022) for IDA and marginally significant (p=0.100) for NIDA. Milk protein was lower for mothers with elevated inflammation, but not significantly so. There was no significant interaction between anemia and elevated inflammation as predictors of milk protein.

Of covariates, time postpartum was a significant positive predictor for fat (p=0.013) and a significant negative predictor for lactose (p=0.020) and protein (p<0.001). Parity was a significant negative predictor for fat (p=0.033), and breastfeeding frequency was a significant negative predictor for protein (p=0.003) and a marginal positive predictor for lactose (p=0.051). None of these variables had significant interactions with anemia in predicting milk nutrients.

Table 4 summarizes overall model predictions for the milk nutrient levels associated with IDA/NIDA/inflammation. In short, after adjusting for covariates, the models predicted that mothers with severe iron deficiency (IDA) produced milk with higher protein and lower fat concentrations, unless they were undergoing acute inflammation/infection, in which case they produced milk with higher protein and higher fat concentrations. In contrast, mothers with NIDA produced milk with higher protein and higher lactose concentrations, and lower fat concentration, unless they were undergoing elevated inflammation/infection, in which case they produced milk with higher concentrations of all three nutrients.

Table 4.

Predicted changes in milk nutrients accompanying IDA/NIDA.

ANEMIA TYPE PROTEIN FAT LACTOSE
IDA No inflammation ↑ ↓ ns
Inflammation ↑
NIDA No inflammation ↑ ↓ ↑
Inflammation ↑
Open in a new tab

IDA, Iron deficiency anemia; NIDA, Non-iron deficiency anemia; ns, not significant

Regression models predict that mothers with IDA produced milk with higher protein and lower fat concentrations, unless they were undergoing acute inflammation/infection, in which case they produced milk with higher protein and higher fat concentrations. In contrast, mothers with NIDA produced milk with higher protein and higher lactose concentrations, and lower fat concentration, unless they were undergoing elevated inflammation/infection, in which case they produced milk with higher concentrations of all three nutrients.

Arrows indicate direction of association; arrow size indicates statistical significance

Discussion

This study evaluated the association of maternal IDA, NIDA, and inflammation/infection with milk macronutrients. It was hypothesized that IDA and NIDA would be associated with reduced milk fat and lactose with differing magnitudes, while IDA and NIDA would be associated with elevated milk protein. It was further hypothesized that anemia in combination with inflammation would be associated with reduced milk nutrients. As hypothesized, both IDA and NIDA (without inflammation) were associated with reduced fat and elevated protein. The results for lactose were against our expectation: milk lactose did not differ between IDA and non-anemic mothers, and was higher among mothers with NIDA. Also unexpectedly, the presence of inflammation/infection reversed both IDA and NIDA’s associations with milk fat from inverse to positive. These findings suggest that, in northern Kenya, mothers with anemia deliver more protein to infants via milk – significantly so among mothers with IDA compared to non-anemic mothers – while mothers with NIDA may deliver significantly more milk lactose. Mothers with anemia deliver less milk fat than non-anemic mothers; however, this pattern reverses for mothers with concurrent anemia and elevated inflammation, who may deliver more fat – this reversal may be especially noticeable among mothers with IDA but may also be observable with NIDA. It should be cautioned that there is a good possibility that other, unaccounted factors impacted milk nutrients, particularly milk fat; nursing behaviors (other than breastfeeding frequency) or mother’s hydration status (which may be altered in inflammation) may influence these associations.

The findings of the present study suggest that specific aspects of maternal nutrition or conditions such as anemia, micronutrient deficiency, or subclinical inflammation/infection can complicate the association between maternal nutrition and milk macronutrients. This may partially explain some of the differing results in the literature on milk macronutrients in relation to maternal nutrition, since different populations have different nutritional and epidemiological backgrounds. If mothers are generally buffering infants from nutritional stress, then there should be no association between maternal nutrition/health and milk nutrient content. In specific or severe malnutrition, maternal buffering may become too costly, or outright impossible. Severe maternal undernutrition, such as anemia (e.g., IDA) should then be associated with impaired delivery of one or more nutrients to milk. If a milk nutrient(s) can be increased to partially compensate for a shortfall of others, then it would be expected that some components decrease while others increase. It may be that in absence of inflammation/infection, anemic mothers elevate milk protein by increasing mammary protein synthesis to partially buffer the compromise in milk fat. Further, in the presence of inflammation/infection, anemic mothers may elevate both protein and fat in milk. This may mitigate compromises in other milk components (e.g., micronutrients or immune factors, both of which were reduced in maternal anemia in El-Farrash et al., 2012 and França et al., 2013) or improve infants’ prospects for surviving an infectious disease. It is possible that anemia has opposite effects on milk fat in the absence (lower) and presence (higher) of inflammation because inflammation alters physiology in some way that makes it easier to deliver fat to milk, or because the costs of elevating milk fat are not outweighed by the benefit in anemia alone, but are in the combination of anemia and elevated inflammation. These hypotheses deserve further investigation to clarify how mothers may buffer infant nutrition against nutritional and other stresses in harsh environments. Mothers elsewhere may buffer milk in different ways, reflecting their nutritional and epidemiological histories of different populations.

Milk protein

Anemia-associated increases in milk protein and decreases in milk fat in this study are congruent with previous research from Brazil (França et al., 2013; in transitional and mature milk but not in colostrum). It is possible that elevated protein may buffer milk nutritional value to some degree against impaired delivery of fat during anemia. Mammary tissues may be able to more readily adjust protein via de novo synthesis during maternal anemia while the adjustment of fat may be more difficult because it would require increased lipid concentrations in the blood and/or enhanced mammary uptake from the blood.

Milk lactose

To our knowledge, no previous research has evaluated the association of maternal anemia with lactose, except for our recent study (Corbitt et al., 2018a, 2018b), which found a non-significant association. The existing literature tends to suggest that lactose levels may be relatively constant in human milk (Aumeistere et al., 2017; Ballard & Morrow, 2013; Lönnerdal, 1986) although contrary findings are accumulating, indicating that lactose has relationships with a host of maternal factors, including maternal age, BMI, weaning, and the resumption of menses (Chang et al., 2015; Dewey, Finly, & Lönnerdal, 1984; Lubetsky, Sever, Mimouni, & Mandel, 2015; Nommsen et al., 1991). The present study joins this small but growing list of studies with the finding that maternal NIDA has an association with increased milk lactose. Our finding of no association for IDA is congruent with constant lactose, which is suggestive of buffering of milk lactose. Increased lactose for NIDA could suggest compensation; this matches our prediction that buffering may involve compensatory increase in one nutrient in association with anemia, while it contradicts our further prediction that compensation would be more likely for protein and not lactose.

Milk fat

The interaction between anemia and inflammation—an increase in milk fat when both are present, when either alone is associated with a decrease in milk fat—may provide infants with additional energy and fat-soluble vitamins to mount an immune response. To the extent that maternal inflammation results from infectious disease, which could be transmitted to the infant, an increase in milk fat in response to inflammation may facilitate a robust immune response, which can be quite energetically and nutritionally expensive (Muehlenbein, Hirschtick, Bonner, & Swartz, 2010; Urlacher et al., 2018; Stephensen, 2001). Given how dangerous the combination of undernutrition and infectious disease is to infants (Katona & Katona-Apte, 2008; Scrimshaw, Taylor, & Gordon, 1968; Pelletier, Frongillo, Schroeder, & Habicht, 1995), the combination in mothers of anemia (which could mean impaired nutrient delivery to milk) and inflammation (which could mean risk for transmission of an infectious agent to the child) may represent a situation that warrants increased delivery of milk fat to the infant, even at high physiological cost to mothers. Thus, our results paint a picture not only of impaired fat delivery to milk during anemia, and buffering in the form of unchanged or increased lactose and protein delivery, but also enhanced delivery of milk fat to infants who are likely to come into contact with an infectious agent, for whom impaired fat delivery could be more dangerous.

Limitations

The present study suffers some limitations. First, it was based on seemingly healthy mothers without clinical symptoms of infection: all episodes of infectious disease included here were subclinical—evident only as elevated inflammation. It is possible that anemia accompanying subclinical inflammation/infection and anemia accompanying clinically apparent infectious disease have different associations with milk nutrients. If so, our findings may not be generalizable beyond subclinical inflammation. Second, the present study did not have information on milk volume. It is therefore not possible to determine if maternal anemia altered the total amount of nutrients transferred to the infant through breastfeeding. The decreased or increased levels of milk nutrients associated with anemia in this study do not necessarily mean anemia compromises or enhances milk nutrient transfer from the mother to offspring. Third, this study analyzed total protein and lactose in archived milk exposed to long-term storage and repeated freeze-thaw cycles. While the effects of freeze-thaw cycles are insignificant for total protein, they are significant for lactose (García-Lara et al., 2012). It is possible that freeze-thaw effects introduced some biases to the present study, particularly to lactose data. It is, however, unlikely that this strongly influenced the study’s findings. All milk specimens, regardless of the anemia status or type, underwent the same number of freeze-thaw cycles. The nutrient levels found in this study were comparable to published ranges from other human populations (Ballard & Morrow, 2013; Miller et al., 2013). It was assumed that freeze-thaw effects do not hinder research on within-population variation of milk characteristics. Lastly, sample size was small for NIDA and therefore results are tentative.

The findings indicate that associations between milk macronutrients and maternal nutritional or disease stress are complex. The directionality of the associations may depend on the type of anemia, the presence of inflammation, and the milk nutrient. This study showed that distinguishing different types of anemia, namely IDA and NIDA and their interaction with inflammation, can allow a more nuanced understanding of the complex ways maternal anemia and inflammation may influence breast milk characteristics and, potentially, infant health. IDA is widespread among postpartum women as a result of high iron requirements to carry pregnancy to term. Findings from this research therefore would help advance our ecological understanding of human milk variation.

Acknowledgements

The study was partially funded by grants to MF from NSF (BCS #1638167 and BCS #0622358) and the Wenner-Gren Foundation (Grant #7460 & 9278). NPR was funded by an NSF Graduate Research Fellowship during a portion of the research and writing of this article. EB was funded by a Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, P2C HD042828, to the Center for Studies in Demography & Ecology at the University of Washington. Emma Bignall, Savannah Sass, Sabrina Shingleton, Izzy Yabes, and Sabrina Perlman of MSU Anthropology assisted with milk sample preparation and assays. The use of milk specimens was possible via a Material Transfer Agreement between the Biodemography Laboratory of the University of Washington and the Biomarker Laboratory for Anthropological Research. Preliminary findings from pilot and final phases of this study were presented in conferences (Corbitt, Paredes Ruvalcaba, & Fujita, 2018a; Fujita, Brindle, & Lo, 2015; Fujita, Paredes Ruvalcaba, Wander, Corbitt, & Brindle, 2018).

Grant sponsorship: NSF, BCS 0622358 & BCS 1638167 (MF); the Wenner-Gren Foundation, Gr. 7460 & 9278 (MF); Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, P2C HD042828 (EB).

Footnotes

Conflict of Interest: None

References

  1. Alene Addis K., & Mohamed Dohe A (2014). Prevalence of anemia and associated factors among pregnant women in an urban area of eastern Ethiopia. Anemia, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Antelman G, Msamanga GI, Spiegelman D, Urassa EJ, Narh R, Hunter DJ, & Fawzi WW (2000). Nutritional factors and infectious disease contribute to anemia among pregnant women with human immunodeficiency virus in Tanzania. The Journal of Nutrition, 130(8), 1950–1957. [DOI] [PubMed] [Google Scholar]
  3. Aumeistere L, Ciproviča I, Zavadska D, & Ceļmalniece K (2017). Lactose content of breast milk among lactating women in Latvia. 2255-9817. Proceedings from Foodbalt 2017: 11th Baltic conference on food science and technology, Jelgava Accessed at: http://agris.fao.org/agris-search/search.do?recordID=LV2017000292. [Google Scholar]
  4. Ballard O, & Morrow AL (2013). Human milk composition: nutrients and bioactive factors. Pediatric Clinics, 60(1), 49–74. doi: 10.1016/j.pcl.2012.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bachour P, Yafawi R, Jaber F, Choueiri E, & Abdel-Razzak Z (2012). Effects of smoking, mother’s age, body mass index, and parity number on lipid, protein, and secretory immunoglobulin A concentrations of human milk. Breastfeeding Medicine, 7(3), 179–188. [DOI] [PubMed] [Google Scholar]
  6. Bodnar LM, Scanlon KS, Freedman DS, Siega-Riz AM, & Cogswell ME (2001). High prevalence of postpartum anemia among low-income women in the United States. American Journal of Obstetrics and Gynecology, 185(2), 438–443. [DOI] [PubMed] [Google Scholar]
  7. Brindle E, Fujita M, Shofer J, & O’Connor KA (2010). Serum, plasma, and dried blood spot high-sensitivity C-reactive protein enzyme immunoassay for population research. Journal of Immunological Methods, 362(1-2), 112–120. doi: 10.1016/j.jim.2010.09.014 Epub 2010 Sep 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cassat JE, & Skaar EP (2013). Iron in infection and immunity. Cell Host & Microbe, 13(5), 509–519. doi: 10.1016/j.chom.2013.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chang N, Jung JA, Kim H, Jo A, Kang S, Lee SW, & Jung BM (2015). Macronutrient composition of human milk from Korean mothers of full term infants born at 37-42 gestational weeks. Nutrition research and practice, 9(4), 433–438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Christian P (2005). Iron deficiency and anemia causes, consequences, and solutions [Online Power Point Presentation]. Retrieved from: ocw.jhsph.edu/courses/InternationalNutrition/PDFs/Lecture5.pdf.
  11. Corbitt M, Paredes Ruvalcaba N, & Fujita M (2018a). Exploring associations between maternal anemia and breast milk macronutrients using samples from Ariaal women in northern Kenya. [Abstract]. University Undergraduate Research and Arts Forum, Michigan State University; Retrieved from https://urca.msu.edu/files/forums/30/booklet/URPBweb2018.pdf. [Google Scholar]
  12. Corbitt M, Paredes Ruvalcaba N, & Fujita M (2018b). Investing associations between maternal anemia and breast milk macronutrients using samples from Ariaal women in northern Kenya. Manuscript submitted for publication.
  13. Darnton-Hill J, Paragas N, Cavalli-Sforza T (2007). Global perspectives: accelerating progress on preventing and controlling nutritional anemia In Kraemer K & Zimmermann M (Eds.), Nutritional Anemia (359–382). Basel, Switzerland: Sight and Life Press. [Google Scholar]
  14. Dewey KG, Finley DA, & Lönnerdal B (1984). Breast milk volume and composition during late lactation (7-20 months). Journal of Pediatric Gastroenterology and Nutrition, 3(5), 713–720. [DOI] [PubMed] [Google Scholar]
  15. Drakesmith H, & Prentice AM (2012). Hepcidin and the iron-infection axis. Science, 338(6108), 768–72. doi: 10.1126/science.1224577. [DOI] [PubMed] [Google Scholar]
  16. El-Farrash RA, Ismail EA, & Nada AS (2012). Cord blood iron profile and breast milk micronutrients in maternal iron deficiency anemia. Pediatric Blood Cancer, 58(2), 233–238. [DOI] [PubMed] [Google Scholar]
  17. Emmett PM, & Rogers IS (1997). Properties of human milk and their relationship with maternal nutrition. Early Human Development, 49, S7–S28. [DOI] [PubMed] [Google Scholar]
  18. França EL, Silva VA, Volpato RM, Silva PA, Brune MF, & Honorio-França AC (2013). Maternal anemia induces changes in immunological and nutritional components of breast milk. The Journal of Maternal-Fetal & Neonatal Medicine, 26(12), 1223–1227. doi: 10.3109/14767058.2013.776529. [DOI] [PubMed] [Google Scholar]
  19. Fratkin E (2004). Ariaal Pastoralists of Northern Kenya (2nd ed.). Needham Heights, MA: Allyn and Bacon. [Google Scholar]
  20. Fratkin E, & Roth E (1990). Drought and economic differentiation among Ariaal Pastoralists of Kenya. Human Ecology, 18, 385–402. [DOI] [PubMed] [Google Scholar]
  21. Fratkin E, & Roth EA (2005). As pastoralists settle: Social, health, and economic consequences of the pastoral sedentarization in Marsabit District, Kenya. In Series: Studies in human ecology and adaptation (Vol. 1). New York: Kluwer Academic, Plenum Publishers. [Google Scholar]
  22. Fratkin E, & Smith K (1995) Women’s changing economic roles with pastoral sedentarization: Varying strategies in alternative Rendille communities. Human Ecology, 23, 433–454. [Google Scholar]
  23. Freedman B (2001). Cytomegalovirus seropositivity and C-reactive protein have independent and combined predictive value for mortality in patients with angiographically demonstrated coronary artery disease. Circulation, 200, e20–e21. 10.1161/01.CIR.104.5.e20 [DOI] [PubMed] [Google Scholar]
  24. Fujita M (2003). Sedentarization, seasonality, and economic differentiation: maternal diet and health in Ariaal-Rendille communities in northern Kenya. (Master’s thesis). University of Victoria, Victoria, Canada, Department of Anthropology. [Google Scholar]
  25. Fujita M (2008). An epidemiological and evolutionary investigation of mother-offspring Vitamin A transfer. (Doctoral dissertation). University of Washington, Seattle, Department of Anthropology. [Google Scholar]
  26. Fujita M, Brindle E, & Lo Y (2015). Maternal anemia and acute phase reaction during breastfeeding interactively predict milk fat concentration among Ariaal mothers in northern Kenya. American Journal of Human Biology 27(2), 267. [Google Scholar]
  27. Fujita M, Brindle E, Lo Y, Castro P, & Cameroamortegui F (2014). Nutrient intakes associated with elevated serum C-reactive protein concentrations in normal to underweight breastfeeding women in Northern Kenya. American Journal of Human Biology, 26(6), 796–802. doi: 10.1002/ajhb.22600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Fujita M, Lo YJ, & Brindle E (2017). Nutritional, inflammatory, and ecological correlates of maternal retinol allocation to breast milk in agro-pastoral Ariaal communities of northern Kenya. American Journal of Human Biology, 29(4). doi: 10.1002/ajhb.22961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Fujita M, Paredes Ruvalcaba N, Wander K, Corbitt M, & Brindle E (2018). An exploratory study of the adaptive anemia hypothesis: Maternal anemia may compromise or enhance breast milk macronutrient levels depending on the type of anemia, the presence of infection, and the milk component [Presentation]. International Society for Evolution, Medicine, and Public Health annual meeting, Park City, Utah. [Google Scholar]
  30. Fujita M, Roth EA, Nathan MA, & Fratkin E (2004). Sedentism, seasonality and economic status: A multivariate analysis of maternal dietary and health statuses between pastoral and agricultural Ariaal and Rendille communities in northern Kenya. American Journal of Physical Anthropology, 123(3), 277–91. [DOI] [PubMed] [Google Scholar]
  31. Fujita M, Roth EA, Nathan MA, & Fratkin EM (2005). Sedentarization and seasonality: Maternal dietary and health consequences in Ariaal and Rendille communities in northern Kenya In Fratkin EM & Roth EA (Eds.), As Pastoralists Settle: Social, Health, and Economic Consequences of the Pastoral Sedentarization in Marsabit District, Kenya (pp.209–234). In Series: Studies in Human Ecology and Adaptation (Vol. 1). New York: Kluwer Academic, Plenum Publishers. [Google Scholar]
  32. Fujita M, Shell-Duncan B, Ndemwa P, Brindle E, Lo YJ, Kombe Y, & O’Connor K (2011). Vitamin A dynamics in breastmilk and liver stores: a life history perspective. American Journal of Human Biology,23(5), 664–73. doi: 10.1002/ajhb.21195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Fujita M, & Wander K (2017). A test of the optimal iron hypothesis among breastfeeding Ariaal mothers in northern Kenya. American Journal of Physical Anthropology, 164(3), 586–597. doi: 10.1002/ajpa.23299. [DOI] [PubMed] [Google Scholar]
  34. García-Lara NR, Escuder-Vieco D, García-Algar O, De la Cruz J, Lora D, & Pallás-Alonso C (2012). Effect of freezing time on macronutrients and energy content of breastmilk. Breastfeeding Medicine, 7(4), 295–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Gardner AS, Rahman IA, Lai CT, Hepworth A, Trengove N, Hartmann PE, & Geddes DT (2017). Changes in fatty acid composition of human milk in response to cold-like symptoms in the lactating mother and infant. Nutrients, 9(9), 1034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Groer M, Davis M, & Steele K (2004). Associations between human milk SIgA and maternal immune, infectious, endocrine, and stress variables. Journal of Human Lactation, 20(2), 153–158. [DOI] [PubMed] [Google Scholar]
  37. Grote V, Verduci E, Scaglioni S, Vecchi F, Contarini G, Giovannini M, & Agostoni C (2015). Breast milk composition and infant nutrient intakes during the first 12 months of life. European journal of clinical nutrition, 70(2), 250. [DOI] [PubMed] [Google Scholar]
  38. Hinde K, Foster AB, Landis LM, Rendina D, Oftedal OT, & Power ML (2013). Daughter dearest: Sex-biased calcium in mother’s milk among rhesus macaques. American Journal of Physical Anthropology, 151(1), 144–150. [DOI] [PubMed] [Google Scholar]
  39. Jelliffe DB, & Jelliffe EP (1978). The volume and composition of human milk in poorly nourished communities. A review. American Journal of Clinical Nutrition, 31(3), 492–515. [DOI] [PubMed] [Google Scholar]
  40. Jurado Rafael L. (1997) Iron, infections, and anemia of inflammation. Clinical Infectious Diseases 25(4), 888–895. [DOI] [PubMed] [Google Scholar]
  41. Katona Peter & Katona-Apte Judit. (2008). The interaction between nutrition and infection. Clinical Infectious Diseases 46(10), 1582–1588. [DOI] [PubMed] [Google Scholar]
  42. Kent S, Weinberg ED, & Stuart-Macadam P (1994). The etiology of the anemia of chronic disease and infection. Journal of Clinical Epidemiology, 47, 23–33. [DOI] [PubMed] [Google Scholar]
  43. Lönnerdal B (1986). Effects of maternal dietary intake on human milk composition. The Journal of Nutrition, 116(4), 499–513. [DOI] [PubMed] [Google Scholar]
  44. Lubetzky R, Sever O, Mimouni FB, & Mandel D (2015). Human milk macronutrients content: effect of advanced maternal age. Breastfeeding Medicine, 10(9), 433–436. [DOI] [PubMed] [Google Scholar]
  45. Lucas A, Gibbs JA, Lyster RJL, & Baum JD (1978). Creamatocrit: simple clinical technique for estimating fat concentration and energy value of human milk. British Medical Journal, 1(6119), 1018–1020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Lynch S (2007). Iron metabolism In Kraemer K & Zimmermann M (Eds.), Nutritional Anemia (pp. 59–76). Basel, Switzerland: Sight and Life Press. [Google Scholar]
  47. Mandel D, Lubetzky R, Dollberg S, Barak S, & Mimouni FB (2005). Fat and energy contents of expressed human breast milk in prolonged lactation. Pediatrics, 116(3), e432–e435. [DOI] [PubMed] [Google Scholar]
  48. Michaelsen KF, Skafte L, Badsberg JH, & Jørgensen M (1990). Variation in macronutrients in human bank milk: influencing factors and implications for human milk banking. Journal of Pediatric Gastroenterology and Nutrition, 11(2), 229–239. [DOI] [PubMed] [Google Scholar]
  49. Miller EM (2010). Maternal hemoglobin depletion in a settled northern Kenyan pastoral population. American Journal of Human Biology, 22, 768–774. [DOI] [PubMed] [Google Scholar]
  50. Miller EM (2016). Reproductive ecology of iron in women. Yearbook of Physical Anthropology, 159:S172–S19. [DOI] [PubMed] [Google Scholar]
  51. Miller JL (2013). Iron deficiency anemia: a common and curable disease. Cold Spring Harbor Perspectives in Medicine, 3(7), a011866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Miller EM, Aiello MO, Fujita M, Hinde K, Milligan L, & Quinn EA (2013). Field and laboratory methods in human milk research. American Journal of Human Biology, 25(1), 1–11. doi: 10.1002/ajhb.22334. [DOI] [PubMed] [Google Scholar]
  53. Mizuno K, Hatsuno M, Aikawa K, Takeichi H, Himi T, Kaneko A, & Itabashi K (2012). Mastitis is associated with IL-6 levels and milk fat globule size in breast milk. Journal of Human Lactation, 28(4), 529–534. doi: 10.1177/0890334412455946. [DOI] [PubMed] [Google Scholar]
  54. Muehlenbein MP, Hirschtick JL, Bonner JZ, & Swartz AM (2010). Toward quantifying the usage costs of human immunity: altered metabolic rates and hormone levels during acute immune activation in men. American Journal of Human Biology, 22(4), 546–556. [DOI] [PubMed] [Google Scholar]
  55. Murase M, Mizuno K, Nishida Y, Mizuno N, Taki M, Yoshizawa M, & Mukai Y (2009). Comparison of creamatocrit and protein concentration in each mammary lobe of the same breast: does the milk composition of each mammary lobe differ in the same breast? Breastfeeding Medicine, 4(4), 189–195. doi: 10.1089/bfm.2008.0124. [DOI] [PubMed] [Google Scholar]
  56. Murray-Kolb LE, & Beard JL (2007). Iron treatment normalizes cognitive functioning in young women. The American Journal of Clinical Nutrition, 85(3), 778–787. [DOI] [PubMed] [Google Scholar]
  57. Nakagomi A, Freedman SB, & Geczy CL (2000). Interferon-gamma and lipopolysaccharide potentiate monocyte tissue factor induction by C-reactive protein: relationship with age, sex, and hormone replacement treatment. Circulation, 101, 1785–1791. [DOI] [PubMed] [Google Scholar]
  58. Nestel P (2002). Adjusting hemoglobin values in program surveys. Washington, DC: International Nutritional Anaemia Consultative Group, ILSI Human Nutrition Institute. [Google Scholar]
  59. Nommsen LA, Lovelady CA, Heinig MJ, Lönnerdal B, & Dewey KG (1991). Determinants of energy, protein, lipid, and lactose concentrations in human milk during the first 12 months of lactation: the DARLING Study. The American Journal of Clinical Nutrition, 53(2), 457–465. [DOI] [PubMed] [Google Scholar]
  60. Pelletier DL, Frongillo EA Jr, Schroeder DG, & Habicht JP (1995). The effects of malnutrition on child mortality in developing countries. Bulletin of the World Health Organization, 73(4), 443. [PMC free article] [PubMed] [Google Scholar]
  61. Picciano MF (2003). Pregnancy and lactation: physiological adjustments, nutritional requirements and the role of dietary supplements. Journal of Nutrition, 133(6), 1997S–2002S. [DOI] [PubMed] [Google Scholar]
  62. Pond CM (1977). The significance of lactation in the evolution of mammals. Evolution, 31(1), 177–199. [DOI] [PubMed] [Google Scholar]
  63. Prentice AM, Goldberg GR, & Prentice A (1994). Body mass index and lactation performance. European Journal of Clinical Nutrition, 48, S78–89. [PubMed] [Google Scholar]
  64. Quinn EA, Largado F, Power M, & Kuzawa CW (2012). Predictors of breast milk macronutrient composition in Filipino mothers. American Journal of Human Biology, 24(4), 533–540. doi: 10.1002/ajhb.22266. [DOI] [PubMed] [Google Scholar]
  65. Quinn EA, Diki Bista K, Childs G (2016). Milk at altitude: Human milk macronutrient composition in a high-altitude adapted population of Tibetans. American Journal of Physical Anthropology, 159(2):233–43. doi: 10.1002/ajpa.22871. [DOI] [PubMed] [Google Scholar]
  66. Rogers LK, & Velten M (2011). Maternal inflammation, growth retardation, and preterm birth: insights into adult cardiovascular disease. Life Sciences, 89(13-14), 417–421. [DOI] [PubMed] [Google Scholar]
  67. Roth EA, Fratkin E (2005). Introduction: The social, health, and economic consequences of pastoral sedentarization in Marsabit District, northern Kenya In Fratkin EM & Roth EA (Eds.), As Pastoralists Settle: Social, Health, and Economic Consequences of the Pastoral Sedentarization in Marsabit District, Kenya (pp.1–28). In Series: Studies in Human Ecology and Adaptation (Vol. 1). New York: Kluwer Academic, Plenum Publishers. [Google Scholar]
  68. Roth EA, Ngugi EN (2005). Female education in a sedentary Ariaal Rendille community: parental decision-making and biosocial pathways In Fratkin EM & Roth EA (Eds.), As Pastoralists Settle: Social, Health, and Economic Consequences of the Pastoral Sedentarization in Marsabit District, Kenya (pp.255–270). In Series: Studies in Human Ecology and Adaptation (Vol. 1). New York: Kluwer Academic, Plenum Publishers. [Google Scholar]
  69. Scrimshaw NS, Taylor CE, & Gordon JE (1968). Interactions of nutrition and infection. [PubMed]
  70. Semba RD, Shah N, Klein RS, Mayer KH, Schuman P, Vlahov D, & Human Immunodeficiency Virus Epidemiology Research Study Group. (2002). Prevalence and cumulative incidence of and risk factors for anemia in a multicenter cohort study of human immunodeficiency virus–infected and–uninfected women. Clinical Infectious Diseases, 34(2), 260–266. [DOI] [PubMed] [Google Scholar]
  71. Shell-Duncan B, & Yung SA (2004). The maternal depletion transition in northern Kenya: the effects of settlement, development and disparity. Social Science & Medicine, 58(12), 2485–2498 [DOI] [PubMed] [Google Scholar]
  72. Stephensen CB (2001). Vitamin A, infection, and immune function. Annual Review of Nutrition, 21(1), 167–192. [DOI] [PubMed] [Google Scholar]
  73. Stone J, Fujita M (2016). Transferrin receptor protein values in dried blood spots stored for nine years demonstrate strong correlation with transferrin receptor in matched sera and with hemoglobin values assessed at the time of blood collection. American Journal of Human Biology, 28(2), 297. [Google Scholar]
  74. Thurnham DI (2014). Interactions between nutrition and immune function: using inflammation biomarkers to interpret micronutrient status. Proceedings of the Nutrition Society, 73(1), 1–8. doi: 10.1017/S0029665113003662. [DOI] [PubMed] [Google Scholar]
  75. Thurnham DI, & Northrop-Clewes CA (2007). Infection and the etiology of anemia In Kraemer K & Zimmermann M (Eds.), Nutritional Anemia (pp. 231–256). Basel, Switzerland: Sight and Life Press. [Google Scholar]
  76. Urlacher SS, Ellison PT, Sugiyama LS, Pontzer H, Eick G, Liebert MA, & Snodgrass JJ (2018). Tradeoffs between immune function and childhood growth among Amazonian forager-horticulturalists. Proceedings of the National Academy of Sciences, 201717522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Wander K, Shell-Duncan B, & McDade TW (2009). Evaluation of iron deficiency as a nutritional adaptation to infectious disease: an evolutionary medicine perspective. American Journal of Human Biology, 21, 172–179. 10.1002/ajhb.20839. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Wang CD, Chu PS, Mellen BG, & Shenai JP (1999). Creamatocrit and the nutrient composition of human milk. Journal of Perinatology, 19(5), 343. [DOI] [PubMed] [Google Scholar]
  79. Weinberg ED (1984). Iron withholding: a defense against infection and neoplasia. Physiological reviews, 64, 65–102. [DOI] [PubMed] [Google Scholar]
  80. World Health Organization. (2008). Worldwide prevalence of anemia 1993-2005: WHO global database on anaemia. de Benoist B, McLean E, Egli I, Cogswell M (Eds). Geneva: World Health Organization; Retrieved from http://apps.who.int/iris/bitstream/handle/10665/43894/9789241596657_eng.pdf?sequence=1 [Google Scholar]
  81. Zavaleta N, Lanata C, Butron B, Peerson JM, Brown KH, & Lönnerdal B (1995). Effect of acute maternal infection on quantity and composition of breast milk. American Journal of Clinical Nutrition, 62(3), 559–63. [DOI] [PubMed] [Google Scholar]

RESOURCES