Breastfeeding Duration Is Inversely Associated with Postpartum Allostatic Load: A Possible Mechanism for Improved Maternal Health

Abstract

Background:

Allostatic load, a multisystem composite measure of chronic stress reflecting the cumulative wear and tear on the body, has utility in explaining maternal and child health disparities and predicting future health when measured during the peripartum period. Research using cross-sectional data has demonstrated an inverse association between concurrent breastfeeding status and maternal postpartum allostatic load. However, the relationship between breastfeeding duration and postpartum allostatic load has not been determined.

Objective:

This study aimed to examine the association between breastfeeding duration and postpartum allostatic load using prospective data, and to compare the association by race/ethnicity to better understand etiologies of health inequities.

Materials and Methods:

A secondary analysis of a data sample of 1,791 postpartum women from the Community Child Health Research Network was conducted. Multiple linear regression tested the association between and breastfeeding duration and allostatic load (range 0–10, calculated from 10 biomarkers) at 6 and 12 months postpartum, adjusting for race/ethnicity, maternal age, education, poverty level, study center, smoking, marital status, and prepregnancy body mass index. Stratified analyses examined differences in associations by race/ethnicity.

Results:

Breastfeeding duration was inversely associated with maternal allostatic load. In adjusted models, mothers who breastfed ≥6 months had 0.41 points lower allostatic load at 6 months (95% CI: −0.71 to −0.11) and 0.36 points lower allostatic load at 12 months postpartum (95% CI: −0.69 to −0.036) compared with mothers who never breastfed. Effect sizes varied in strength and significance across race/ethnic groups.

Conclusions:

Study findings suggest that extended breastfeeding for at least 6 months is linked with reduced chronic stress load in mothers during the first year postpartum. The inconsistency of findings across race/ethnic groups signals the importance of adjusting for prepregnancy allostatic load in future studies.

Introduction

Within the field of racial disparities in maternal and child health, there is growing interest in using allostatic load, a multisystem composite measure of chronic stress, to capture and analyze the impact of accumulative stress on maternal and child health.^1–4 Allostatic load represents the wear and tear on the systems of the body and consequent susceptibility to illness, resulting from excessive triggering and demands on the body's adaptive response to environmental challenges, otherwise known as allostasis.⁵ While there is no standard and consistent measure of allostatic load, researchers have commonly operationalized allostatic load by combining biomarkers reflecting the functional status of cardiovascular, metabolic, immune, and neuroendocrine systems into a score or index, with higher scores indicating greater vulnerability to illness.^6–8

Studies have found higher allostatic load during the peripartum period among Black women compared with White and Hispanic women,^2,9 explained by socioeconomic disadvantage, greater exposures to lifetime stress, and living in neighborhoods with fewer resources.² Furthermore, women with higher allostatic load are more likely to experience preeclampsia¹⁰ and have low-birthweight infants.^9,10 Understanding racial and ethnic disparities in allostatic load during the peripartum period illuminates the impact of differential exposures to chronic stress on maternal health and draws attention to conditions surrounding fetal programming and infant development contributing to the health of future generations.

The peripartum period presents a critical time for resetting health, when protective factors such as breastfeeding can improve the trajectory of health for mothers and children. Breastfeeding-induced hormonal changes in mothers, including increases in oxytocin and prolactin, have been shown to have downregulating effects on hypothalamic–pituitary–adrenal axis activity to support postpartum recovery from the physiological stress of pregnancy and delivery.¹¹ In addition, research has suggested many long-term benefits of breastfeeding on maternal health, which are magnified with longer breastfeeding duration, including reduced risk of developing breast and ovarian cancers,¹² hypertension,¹³ diabetes,^13,14 and metabolic syndrome.¹⁵

Given the many demonstrated benefits of breastfeeding to women's health,¹² there is limited research on whether breastfeeding interacts with chronic stress (i.e., allostatic load from chronic accommodation to environmental stressors) pathways to improve maternal health. Our previous study of women within 2 years postpartum derived from the National Health and Nutrition Examination Survey (NHANES) found a significant inverse relationship between concurrent breastfeeding status and allostatic load.¹⁶

The current study is a secondary analysis of a prospective cohort aimed to address the limitations in our previous study, the inability to determine temporal relationship and failure to capture breastfeeding duration. In this study, we use longitudinal data to examine the prospective relationship between breastfeeding duration and allostatic load at 6 and 12 months postpartum. We hypothesized longer duration of breastfeeding would be associated with lower allostatic load at 6 and 12 months due to the long-term benefits of breastfeeding. In addition, we aimed to examine associations between breastfeeding duration and allostatic load by race/ethnicity, to better elucidate breastfeeding's influence within the context of racial/ethnic health inequities and patterns of heightened life-course weathering of stress among Black women.³ We hypothesized an inverse relationship among all racial/ethnic groups.

Materials and Methods

Study design

This study is a secondary analysis of longitudinal data generated by the Community and Child Health Network (CCHN). CCHN was a multisite prospective cohort study that recruited 2,510 mothers between 2008 and 2010 around the birth of their child at hospitals and clinics and followed them for 2 years (∼1, 6, 12, and 24 months postpartum).^2,17 To support the prospective nature of this study, predictor variables (i.e., breastfeeding duration and covariates) were chosen to capture exposure before the outcome variable (i.e., allostatic load at 6 and 12 months postpartum).

Data for predictor variables were derived from hospital charts at the time of childbirth (T0), and questionnaires from 1 month (T1) and 6 months (T2) postpartum. The outcome variable allostatic load was derived from clinical and biological data collected at 6 months (T2) and 12 months (T3) postpartum, and incorporated 10 values: body mass index (BMI), ratio of waist to hip circumference, systolic and diastolic blood pressure, pulse, diurnal salivary cortisol, and blood spot measures of high-sensitive C-reactive protein (HS-CRP), hemoglobin A1c (HbA1c), high-density lipoprotein (HDL), low-density lipoprotein, and total cholesterol.

CCHN research instruments were designed based on principles of Community-Based Participatory Research (CBPR) reflecting a close collaboration between community representatives and academic researchers.^1,17 Community representatives collaborated with academic researchers in CCHN working groups to design questionnaires, nominating measures for use and providing input on the appropriateness of questions, scales, and wording for engaging with their communities.^1,17 Community members also collaborated with academic partners in leading research training sessions and received rigorous training so that they could be involved in most of the data collection.^1,17 The review boards at each of the five CCHN participating sites approved the study protocol.

Study population

CCHN study sites included five catchment areas in urban (Baltimore, MD; Washington, DC; Los Angeles County, CA), suburban (Lake County, IL), and rural areas (seven counties in eastern North Carolina) with sufficient diversity to allow for analyses of racial/ethnic disparities and socioeconomic gradients.^1,2,17 Participating mothers were 18–40 years old, self-identified as “Black or African American,” “Hispanic or Latina,” and/or “White,” and gave birth to a live child of ≥20 weeks gestational age who was third or lower in birth order. Additional eligibility criteria are detailed in prior research studies.^2,17 Drawing from a total of 2,510 participating mothers, our study sample included 1,791 mothers with data on allostatic load at T2 or T3, and who did not become pregnant during the first year postpartum since pregnancy could bias biomarker values.¹⁸

Assessment of breastfeeding duration

Data on breastfeeding duration were self-reported through administered questionnaires at 1 and 6 months postpartum. At both time points, participants were asked if they had ever breastfed. If affirmative, participants were asked follow-up questions about current breastfeeding status. Participants were categorized as “never breastfed” if they reported that they had never breastfed and “≥6 months” if they reported they were still breastfeeding at 6 months. If mothers answered that they had breastfed but were not still breastfeeding at the time of the survey, they provided information about the length of time they had breastfed, and were accordingly categorized as breastfed “<1 month” or “≥1 to <6 months.” The four breastfeeding duration categories were created to best capture the distribution of data and to follow commonly reported thresholds of breastfeeding.^19,20

Assessment of allostatic load

Allostatic load was a composite measure of 10 biomarkers derived from clinical and biological data collected at 6 and 12 months postpartum. The clinical threshold method of calculating allostatic load^2,10 was used to increase the clinical relevance of the study. Following other analyses of CCHN data,^2,10 we assigned one point for each of 10 biomarkers that exceeded the clinical risk threshold and summed them to create an allostatic load index with range of 0–10 points. The following biomarkers and clinical cutoffs were used: (1) BMI, ≥30 kg/m²; (2) waist/hip ratio, ≥0.85; (3) systolic blood pressure, ≥125 mmHg; (4) diastolic blood pressure, ≥80 mmHg; (5) pulse, ≥100 beats per minute; (6) HS-CRP, ≥3 mg/L; (7) HbA1c, ≥5.4%; (8) HDL, ≤40 mg/dL; (9) total cholesterol/HDL ratio, ≥5.9; and (10) diurnal cortisol slope, ≥−0.01.²

Assessment of covariates

Data on race/ethnicity were self-reported at the time of recruitment (T0) based on mothers' choice of primary racial/ethnic identification as “Black or African American,” “White,” or “Hispanic.” Hospital chart reviews from T0 were used to collect the following covariate data: maternal age, parity (first child/second child/≥2 previous children), birth method (Cesarean/vaginal), preterm birth (yes/no), delivery of low-birthweight infant (yes/no), and prepregnancy weight and height, which we converted to BMI (kg/m²). Data on household income (converted to % federal poverty level), maternal education (<high school/high school graduate/some college/≥4-year college degree), marital status (married/not married), current smoking status (yes/no), employment (working full time/working part time/unemployed/on paid leave/on unpaid leave/full time homemaker/student/disabled), and nativity (born in the United States/outside the United States) were self-reported through questionnaires at T1.

Exposure to stressors was a composite index of seven validated self-reported measures of stress exposure (i.e., discrimination, interpersonal violence, pregnancy stress, life events, financial stress, chronic stress, perceived stress) collected at T1and T2 following previously implemented methods of assigning one point to those who fell in the upper quartile for each stressor and summing them for an overall score (range 0–7).²

Statistical analyses

STATA 17.0²¹ was used for all statistical analyses, and statistical significance was determined at p < 0.05. We conducted preliminary bivariate analyses to examine the relationship between all covariates of interest and breastfeeding duration categories. Skewness tests were first conducted to check normality of variable distributions. The Kruskal–Wallis nonparametric one-way analysis of variance (ANOVA) was used to examine differences by breastfeeding duration categories for continuous covariates (i.e., maternal age, prepregnancy BMI, exposure to stressors), which were not normally distributed. Chi-square was used to examine differences by breastfeeding duration for the remaining covariates which were categorical. The bivariate associations between breastfeeding duration and allostatic load at T2 and T3, as well as between breastfeeding duration and each of the 10 biomarkers used in calculating allostatic load, were examined using Kruskal–Wallis nonparametric one-way ANOVA followed by the Dunn's test for paired comparisons.

Multiple linear regression models tested the association between breastfeeding duration and allostatic load at T2 and T3 and residuals were checked to ensure assumptions of normality were met (skewness = 0.36 at T2 and 0.54 at T3). Covariates were selected for inclusion in the final models based on a-priori determination, augmented backward elimination, and bivariate analyses with the goal of creating a parsimonious model integrating the most important variables. Selection of race/ethnicity, maternal age, and indicators of socioeconomic status (i.e., maternal education, and poverty status) was based on established relationships with both breastfeeding^22,23 and allostatic load.^6,24,25 Augmented backward elimination,²⁶ with significance threshold set at α = 0.2 and change-in-estimate threshold set at τ = 0.05, was used to select study center and prepregnancy BMI, which were significant in both T2 and T3 models. Bivariate analyses were used to consider smoking, marital status, and preterm birth, which were significant in one but not both T2 and T3 models using augmented backward elimination.

We selected smoking and marital status as covariates since they were significantly and independently associated with breastfeeding and allostatic load. The final regression models included race/ethnicity, maternal age, maternal education, poverty status, study center, prepregnancy BMI, smoking, and marital status as covariates. For exploration of associations for each racial/ethnic group, analyses were stratified by race/ethnicity. A-priori power analysis using G*Power 3.1²⁷ indicated the need for a sample size of ≥550 to detect a small effect size (Cohen's f² = 0.02)²⁸ with ≥80% power and 18 predictor variables and a sample size of ≥78 to detect a medium effect size (Cohen's f² = 0.15).²⁸

Multiple linear regression models included participants with complete data on allostatic load at T2 (n = 1,417) and T3 (n = 1,467). Data were also complete for race/ethnicity, maternal age, poverty status, marital status, and study center. Given that examination of missing covariate data showed missing at random patterns, multiple imputation and pooled estimates from 10 imputations were used to account for missing covariate data in adjusted regression analyses.²⁹

Results

Descriptive statistics and bivariate associations

In the sample of 1,791 mothers in the first year postpartum, of those with complete data on breastfeeding (n = 1,624), 28% never breastfed their infant, 21% breastfed <1 month, 27% breastfed ≥1 to <6 months, and 24% breastfed ≥6 months (Table 1). Black women comprised more than half (53%) of the analytical sample and among Black mothers with breastfeeding data, 39% reported never breastfeeding and 13% had breastfed for ≥6 months. This contrasted with a lower proportion of White (21%) and Hispanic (9%) mothers who had never breastfed, as well as a higher proportion of White (40%) and Hispanic (36%) mothers who breastfed ≥6 months. Across the sample, mothers who breastfed ≥6 months tended to be older and in the highest income category (i.e., >200% Federal Poverty Level). Compared with those who never breastfed, the proportion of mothers who breastfed ≥6 months was significantly higher for those who had a 4-year college degree, were born outside the United States, were married, were not smoking, did not give birth preterm, and did not have a low-birthweight infant (Table 1).

Table 1.

Participant Characteristics by Breastfeeding Duration Category in the Community Child Health Research Network (N = 1791)

	Total (N = 1,791)	Never breastfed (n = 455)	Breastfed <1 month (n = 344)	Breastfed ≥1 to <6 months (n = 434)	Breastfed ≥6 months (n = 391)	Missing breastfeeding (n = 167)	p ^a
Age (years), median (mean ± SD)	24.61 (25.93 ± 5.75)	22.35 (23.77 ± 4.55)	22.52 (24.41 ± 5.44)	24.86 (26.05 ± 5.59)	28.33 (28.67 ± 5.87)	28.19 (28.24 ± 5.98)	0.0001
Prepregnancy BMI, median (mean ± SD)	25.78 (27.50 ± 7.35)	26.16 (28.04 ± 8.16)	26.17 (27.62 ± 7.27)	26.52 (27.90 ± 7.05)	25.45 (26.73 ± 6.80)	24.03 (26.85 ± 7.36)	0.1449
Stress exposure (0–7), median (mean ± SD)	2 (1.96 ± 1.71)	2 (1.99 ± 1.77)	2 (2.04 ± 1.73)	2 (1.96 ± 1.64)	1.75 (1.82 ± 1.63)	1.4 (2.08 ± 1.91)	0.3985
Race/ethnicity, n (%)							<0.0001
White	407 (22.72)	75 (16.48)	69 (20.06)	78 (17.97)	143 (36.57)	42 (25.15)
Black	957 (53.43)	349 (76.70)	186 (54.07)	238 (54.84)	117 (29.92)	67 (40.12)
Hispanic	427 (23.84)	31 (6.81)	89 (25.87)	118 (27.19)	131 (33.50)	58 (34.73)
Education, n (%)							<0.0001
<high school	316 (18.06)	105 (23.39)	60 (17.70)	70 (16.79)	52 (13.58)	29 (17.90)
High school graduate	760 (43.43)	250 (55.68)	171 (50.44)	162 (38.85)	124 (32.38)	53 (32.72)
Some college but not 4-year degree	405 (23.14)	75 (16.70)	81 (23.89)	132 (31.65)	84 (21.93)	33 (20.37)
4-year degree or higher	269 (15.37)	19 (4.23)	27 (7.96)	53 (12.71)	123 (32.11)	47 (29.01)
Federal poverty level, n (%)							<0.0001
≤100%	758 (42.32)	267 (58.68)	163 (47.38)	166 (38.25)	108 (27.62)	54 (32.34)
>100% to ≤200%	492 (27.47)	116 (25.49)	99 (28.78)	132 (30.41)	99 (25.32)	46 (27.54)
>200%	541 (30.21)	72 (15.82)	82 (23.84)	136 (31.34)	184 (47.06)	67 (40.12)
Study center, n (%)							<0.0001
Baltimore	412 (23.00)	173 (38.02)	64 (18.60)	71 (16.36)	67 (17.14)	37 (22.16)
Chicago suburb	441 (24.62)	70 (15.38)	101 (29.36)	119 (27.42)	115 (29.41)	36 (21.56)
Los Angeles	244 (13.62)	14 (3.08)	35 (10.17)	61 (14.06)	76 (19.44)	58 (34.73)
North Carolina counties	323 (18.03)	127 (27.91)	71 (20.64)	78 (17.97)	42 (10.74)	5 (2.99)
Washington, DC	371 (20.71)	71 (15.60)	73 (21.22)	105 (24.19)	91 (23.27)	31 (18.56)
Mother employment status, n (%)							<0.0001
Working full time	213 (11.97)	44 (9.67)	45 (13.16)	68 (15.78)	35 (9.09)	21 (12.65)
Working part time	181 (10.17)	40 (8.79)	34 (9.94)	41 (9.51)	44 (11.43)	22 (13.25)
Unemployed	623 (35.02)	208 (45.71)	126 (36.84)	146 (33.87)	102 (26.49)	41 (24.70)
On paid leave	187 (10.51)	27 (5.93)	29 (8.48)	44 (10.21)	66 (17.14)	21 (12.65)
On unpaid leave	180 (10.12)	45 (9.89)	26 (7.60)	44 (10.21)	45 (11.69)	20 (12.05)
Full-time homemaker	259 (14.56)	53 (11.65)	47 (13.74)	61 (14.15)	71 (18.44)	27 (16.27)
Student	86 (4.83)	24 (5.27)	26 (7.60)	16 (3.71)	13 (3.38)	7 (4.22)
Disabled or other	50 (2.81)	14 (3.08)	9 (2.63)	11 (2.55)	9 (2.34)	7 (4.22)
Nativity, n (%)							<0.0001
Born in United States	1,426 (79.66)	435 (95.81)	294 (85.47)	331 (76.27)	251 (63.19)	115 (68.86)
Born outside of United States	364 (20.34)	19 (4.19)	50 (14.53)	103 (23.73)	140 (35.81)	52 (31.14)
Married, n (%)							<0.0001
No	1,210 (67.56)	400 (87.91)	261 (75.87)	296 (68.20)	180 (46.04)	73 (43.71)
Yes	581 (32.44)	55 (12.09)	83 (24.13)	138 (31.80)	211 (53.96)	94 (56.29)
Parity, n (%)							0.323
First child	624 (37.10)	156 (38.90)	134 (41.10)	140 (34.06)	127 (33.16)	67 (41.61)
One previous child	623 (37.04)	145 (36.16)	115 (35.28)	158 (38.44)	151 (39.43)	54 (33.54)
≥2 previous children	435 (25.86)	100 (24.94)	77 (23.62)	113 (27.49)	105 (27.42)	40 (24.84)
Delivery type, n (%)							0.905
Vaginal	1,071 (63.15)	262 (63.75)	203 (61.89)	255 (62.04)	246 (63.90)	105 (65.22)
Cesarean	625 (36.85)	149 (36.25)	125 (38.11)	156 (37.96)	139 (36.10)	56 (34.78)
Preterm birth, n (%)							0.048
No	1,548 (86.72)	397 (87.64)	290 (84.80)	359 (83.10)	349 (89.26)	153 (91.62)
Yes	237 (13.28)	56 (12.36)	52 (15.20)	73 (16.90)	42 (10.74)	14 (8.38)
Low-birthweight infant, n (%)							0.002
No	1,316 (89.95)	307 (87.97)	257 (91.13)	296 (86.05)	316 (94.33)	140 (91.50)
Yes	147 (10.05)	42 (12.03)	25 (8.87)	48 (13.95)	19 (5.67)	13 (8.50)
Smoking, n (%)							<0.0001
No	1,489 (83.32)	337 (74.07)	268 (78.36)	370 (85.25)	364 (93.33)	150 (90.36)
Yes	298 (16.68)	118 (25.93)	74 (21.64)	64 (14.75)	26 (6.67)	16 (9.64)

p-Values based on Kruskal–Wallis nonparametric one-way ANOVA for maternal age, prepregnancy BMI, and stressor exposure, and Chi-Square for all other variables, and excludes missing breastfeeding data.

ANOVA, analysis of variance; BMI, body mass index.

Median and mean biomarker values of the total sample fell below the clinical risk threshold for all biomarkers, except HS-CRP, which had mean values higher than the clinical threshold of ≥3 mg/L at T2 and T3 (Table 2). Pairwise comparisons showed that mothers who breastfed ≥6 months had significantly lower median allostatic load, BMI, systolic and diastolic blood pressure, pulse, and a steeper cortisol slope at T2 and T3 than mothers who never breastfed; HS-CRP was significantly lower at T3 but not T2; and HDL levels were significantly higher (indicating better health) at T2 but not T3. From T2 to T3, change in median values of biomarkers and allostatic load was minimal (≤5% change for all values) (Table 2).

Table 2.

Mean and Median Biomarker Values at 6 and 12 Months Postpartum by Breastfeeding Duration Categories in the Community Child Health Research Network (N = 1,791)

Biomarker	Data collection time^a	High-risk clinical value (% participants with high risk)	Total (n = 1,791) Median (mean ± SD)	1 Never breastfed (n = 455) Median (mean ± SD)	2 Breastfed <1 month (n = 344) Median (mean ± SD)	3 Breastfed ≥1 to <6 months (n = 434) Median (mean ± SD)	4 Breastfed ≥6 months (n = 391) Median (Mean ± SD)	p^b Comparison across groups	Paired comparisons with significance^c
BMI (kg/m²)	T2	≥30 (42.36)	28 (29.40 ± 7.68)	29 (30.64 ± 8.49)	29 (30.25 ± 7.65)	28.95 (29.93 ± 7.31)	26.5 (28.05 ± 7.04)	0.0001	1 and 4, 2 and 4, 3 and 4
BMI (kg/m²)	T3	≥30 (42.87)	28.1 (29.02 ± 7.80)	29.9 (31.07 ± 8.95)	28.6 (29.92 ± 7.64)	29 (29.71 ± 7.60)	26.6 (28.09 ± 7.15)	0.0001	1 and 4, 2 and 4, 3 and 4
Waist: hip ratio	T2	≥0.94 (18.31)	0.87 (0.87 ± 0.075)	0.87 (0.87 ± 0.085)	0.88 (0.88 ± 0.086)	0.86 (0.87 ± 0.082)	0.87 (0.86 ± 0.072)	0.2037	2 and 3, 2 and 4
Waist: hip ratio	T3	≥0.94 (18.95)	0.86 (0.86 ± 0.078)	0.87 (0.87 ± 0.086)	0.87 (0.87 ± 0.087)	0.86 (0.86 ± 0.079)	0.86 (0.86 ± 0.076)	0.2193	2 and 3
Systolic blood pressure (mmHg)	T2	≥120 (20.28)	110.67 (111.18 ± 11.20)	112.33 (112.43 ± 12.00)	111.83 (112.65 ± 12.38)	111.33 (111.41 ± 10.91)	107.33 (108.30 ± 11.56)	0.0001	1 and 4, 2 and 4, 3 and 4
Systolic blood pressure (mmHg)	T3	≥120 (20.95)	110.67 (110.93 ± 11.45)	110.67 (111.83 ± 12.78)	111.67 (112.49 ± 12.23)	111 (111.09 ± 11.89)	109.33 (109.65 ± 10.67)	0.0508	1 and 4, 2 and 4
Diastolic blood pressure (mmHg)	T2	≥80 (24.79)	73 (73.66 ± 9.63)	74.33 (74.86 ± 10.13)	72.67 (73.95 ± 9.66)	73.67 (74.40 ± 10.19)	71 (72.08 ± 9.25)	0.0007	1 and 4, 2 and 4, 3 and 4
Diastolic blood pressure (mmHg)	T3	≥80 (22.52)	73 (73.17 ± 9.55)	74.33 (74.23 ± 9.79)	72.17 (72.97 ± 9.79)	74 (73.39 ± 10.67)	70.83 (71.82 ± 9.09)	0.0052	1 and 2, 1 and 4, 3 and 4
Pulse (beats per minute)	T2	≥91 (9.09)	76.33 (76.38 ± 10.37)	78.67 (79.36 ± 9.66)	75.17 (76.72 ± 11.45)	75.83 (76.13 ± 10.90)	74 (74.75 ± 9.88)	0.0001	1 and 2, 1 and 3, 1 and 4, 2 and 4, 3 and 4
Pulse (beats per minute)	T3	≥91 (9.05)	76.67 (76.66 ± 9.93)	79.33 (78.71 ± 9.72)	75.33 (76.51 ± 10.84)	75.67 (76.38 ± 10.50)	75.67 (76.23 ± 10.12)	0.0009	1 and 2, 1 and 3, 1 and 4
High-sensitive C-reactive protein (mg/L)	T2	≥3 (47.26)	2.7 (4.01 ± 3.89)	2.8 (4.04 ± 3.59)	3.3 (4.27 ± 3.57)	2.6 (4.02 ± 3.95)	2.35 (3.65 ± 3.68)	0.0786	2 and 4
High-sensitive C-reactive protein (mg/L)	T3	≥3 (46.57)	2.6 (4.05 ± 4.16)	3 (4.49 ± 4.29)	2.7 (4.35 ± 4.21)	3.25 (4.59 ± 4.42)	1.9 (3.55 ± 4.12)	0.0012	1 and 4, 2 and 4, 3 and 4
Hemoglobin A1c (%)	T2	≥5.4 (44.26)	5.3 (5.29 ± 0.64)	5.2 (5.30 ± 0.74)	5.3 (5.39 ± 0.85)	5.3 (5.43 ± 0.83)	5.2 (5.24 ± 0.53)	0.0012	1 and 2, 1 and 3, 2 and 4, 3 and 4
Hemoglobin A1c (%)	T3	≥5.4 (44.28)	5.3 (5.31 ± 0.77)	5.3 (5.38 ± 0.90)	5.3 (5.42 ± 1.05)	5.3 (5.40 ± 0.95)	5.2 (5.25 ± 0.59)	0.3766	n.s.
HDL cholesterol (mg/dL)	T2	≤40 (44.59)	43 (44.88 ± 14.66)	41 (42.72 ± 13.23)	42 (44.96 ± 14.48)	43 (45.33 ± 14.10)	45 (46.80 ± 15.24)	0.0072	1 and 2, 1 and 3, 1 and 4
HDL cholesterol (mg/dL)	T3	≤40 (43.71)	43 (44.37 ± 13.41)	42 (44.43 ± 13.40)	44 (45.51 ± 13.71)	42 (44.48 ± 13.60)	42 (44.76 ± 14.41)	0.6797	n.s.
Total cholesterol: HDL ratio	T2	≥5.9 (11.99)	3.96 (4.15 ± 1.43)	4.08 (4.26 ± 1.44)	3.96 (4.04 ± 1.41)	3.95 (4.24 ± 1.54)	3.92 (4.14 ± 1.37)	0.3795	1 and 2
Total cholesterol: HDL ratio	T3	≥5.9 (8.92)	3.73 (3.99 ± 1.39)	3.67 (3.92 ± 1.32)	3.57 (3.93 ± 1.42)	3.82 (4.02 ± 1.30)	3.69 (3.96 ± 1.36)	0.4662	n.s.
Cortisol slope	T2	≥−0.01 (25.06)	−0.022 (−0.022 ± 0.026)	−0.021 (−0.020 ± 0.029)	−0.019 (−0.018 ± 0.029)	−0.020 (−0.018 ± 0.027)	−0.026 (−0.028 ± 0.021)	0.0001	1 and 4, 2 and 4, 3 and 4
Cortisol slope	T3	≥−0.01 (30.55)	−0.021 (−0.023 ± 0.024)	−0.013 (−0.017 ± 0.022)	−0.021 (−0.020 ± 0.026)	−0.019 (−0.023 ± 0.022)	−0.026 (−0.026 ± 0.024)	0.0001	1 and 2, 1 and 3, 1 and 4, 2 and 4, 3 and 4
Allostatic load (0–10 range)	T2	NA	2.86 (2.83 ± 1.87)	3 (3.17 ± 1.98)	3 (3.17 ± 1.82)	3 (3.02 ± 1.92)	2 (2.37 ± 1.87)	0.0001	1 and 4, 2 and 4, 3 and 4
Allostatic load (0–10 range)	T3	NA	2.86 (2.76 ± 1.83)	3 (3.27 ± 2.02)	3 (2.95 ± 1.97)	3 (2.99 ± 1.94)	2 (2.46 ± 1.81)	0.0001	1 and 2, 1 and 4, 2 and 4, 3 and 4

T2 indicates 6 months postpartum; T3 indicates 12 months postpartum.

p-value based on Kruskal–Wallis nonparametric one-way ANOVA comparison of medians.

Significance established at p < 0.05 based on the Dunn's nonparametric pairwise comparison test.

ANOVA, analysis of variance; HDL, high-density lipoprotein; NA, not applicable.

Primary analysis results

Multiple linear regression models indicated a significant inverse association between breastfeeding duration and allostatic load at T2 and T3 (Table 3). Mothers who breastfed ≥6 months had 0.41 points lower allostatic load at T2 (95% CI: −0.71 to −0.11) and 0.36 points lower allostatic load at T3 (95% CI: −0.69 to −0.036) compared with mothers who never breastfed, while controlling for race/ethnicity, maternal age, education, poverty level, study center, smoking, marital status, and prepregnancy BMI (Table 3). When compared with mothers who breastfed <1 month, allostatic load for mothers who breastfed ≥6 months was lower by 0.59 points at T2 (95% CI: −0.88 to −0.29) and 0.24 points (not significant) at T3 (95% CI: −0.57 to 0.096). Tests of interaction by race/ethnicity did not reach statistical significance (p > 0.05 for all interaction terms).

Table 3.

Final Multiple Linear Regression Models of Association Between Breastfeeding Duration and Allostatic Load at 6 Months Postpartum (T2) (n = 1,417) and 12 Months Postpartum (T3) (n = 1,467) in the Community Child Health Research Network

T2^a allostatic load	Adjusted β^b	Standard error	p	95% Confidence interval	T3^c allostatic load	Adjusted β	Standard error	p	95% Confidence interval
Breastfeeding duration
Never breastfed	ref	ref	ref	Ref		ref	ref	ref	ref
<1 month	0.17	0.14	0.23	−0.11 to 0.46		−0.12	0.15	0.41	−0.42 to 0.17
≥1 to <6 months	−0.012	0.14	0.93	−0.28 to 0.26		−0.11	0.15	0.46	−0.40 to 0.18
≥6 months	−0.41	0.15	0.007	−0.71 to −0.11		−0.36	0.16	0.030	−0.69 to −0.036
Race/ethnicity
White	ref	ref	ref	Ref		ref	ref	ref	ref
Black	0.48	0.16	0.003	0.17 to 0.79		0.54	0.16	0.001	0.23 to 0.85
Hispanic	0.38	0.17	0.026	0.045 to 0.72		0.58	0.17	0.001	0.24 to 0.92
Maternal age (years)	0.029	0.011	0.009	0.0072 to 0.051		0.013	0.008	0.23	−0.0082 to 0.035
Maternal education
<High school	ref	ref	ref	Ref		ref	ref	ref	ref
High school graduate	−0.062	0.14	0.65	−0.33 to 0.21		−0.074	0.13	0.58	−0.34 to 0.19
Some college	0.049	0.17	0.77	−0.28 to 0.37		−0.0051	0.17	0.98	−0.33 to 0.32
≥4-year college degree	−0.26	0.22	0.24	−0.69 to 0.17		−0.28	0.22	0.21	−0.72 to 0.16
Federal poverty level
≤00%	ref	ref	ref	Ref	ref	ref	ref	ref	ref
>100% to ≤200%	−0.12	0.12	0.31	−0.36 to 0.11		−0.081	0.12	0.51	−0.32 to 0.16
>200%	−0.21	0.16	0.18	−0.52 to 0.10		−0.13	0.16	0.41	−0.45 to 0.18
Study center
Baltimore	0.47	0.20	0.018	0.082 to 0.86		0.46	0.18	0.010	0.11 to 0.82
Chicago suburb	0.31	0.17	0.074	−0.030 to 0.65		−0.038	0.16	0.815	−0.35 to 0.28
Los Angeles	ref	ref	ref	Ref		ref	ref	ref	ref
North Carolina counties	0.49	0.20	0.014	0.11 to 0.88		0.035	0.19	0.85	−0.34 to 0.41
Washington, DC	0.52	0.19	0.005	0.16 to 0.89		0.12	0.17	0.47	−0.22 to 0.46
Smoking
No	ref	ref	ref	Ref		ref	ref	ref	ref
Yes	0.11	0.14	0.44	−0.16 to 0.38		0.33	0.13	0.012	0.073 to 0.60
Marital status
Not married	ref	ref	ref	Ref		ref	ref	ref	ref
Married	−0.15	0.14	0.28	−0.41 to 0.12		−0.0048	0.13	0.97	−0.26 to 0.25
Prepregnancy BMI	0.093	0.0083	<0.0001	0.077 to 0.11		0.10	0.0083	<0.0001	0.086 to 0.12
Constant	−0.91	0.40	0.024	−1.69 to −0.12		−0.59	0.39	0.13	−1.36 to 0.17

T2 = 6 months postpartum.

Adjusting for all other independent variables in the model.

T3 = 12 months postpartum.

BMI, body mass index; ref, reference category.

Results stratified by race/ethnicity

Stratified analyses by race/ethnicity showed that while many biomarkers were significantly improved by longer breastfeeding duration for White mothers, only a few were significantly improved for Black and Hispanic mothers (Supplementary Table S1).

Among Black mothers, median values of HbA1c at T2 and T3 no longer indicated high risk for mothers who breastfed ≥6 months compared with mothers who breastfed <1 month (5.3% versus 5.4%; clinical risk threshold: ≥5.4%). Black mothers with higher breastfeeding duration at T2 had significantly slower pulse rates and steeper cortisol slopes (indicating better outcome). Among Hispanic mothers, the most notable improvements with breastfeeding duration were with BMI and HS-CRP at T2 and T3, in which median values decreased incrementally with each longer breastfeeding duration category (Supplementary Table S1).

In the race-stratified models, multiple linear regression models indicated a significant inverse association between breastfeeding duration and allostatic load for White mothers at T2 and T3 and Hispanic mothers at T2 (≥6 months versus never breastfeeding). The effect size was largest for Hispanic mothers at T2, who showed 0.91 points lower allostatic load among those who breastfed ≥6 months compared with those who never breastfed (95% CI: −1.67 to −0.14) (Supplementary Tables S2 and S3).

Discussion

This study found that longer duration of breastfeeding (i.e., breastfed ≥6 months) was associated with lower maternal allostatic load during the first year postpartum in adjusted analyses of our full study sample (i.e., not stratified by race). This is the first study that has examined the relationship between breastfeeding duration and allostatic load. Using longitudinal data from CCHN and analyzing the prospective relationship with allostatic load at 6 and 12 months postpartum, this study builds on previous work identifying an inverse relationship between breastfeeding and allostatic load from cross-sectional data.¹⁶

Our previous study found that among mothers who were breastfeeding, allostatic load was 0.36 points lower in the adjusted model, compared with mothers who were not breastfeeding at the time of participating in NHANES.¹⁶ In the current study, a higher effect size (β = −0.41) was found at T2 in the adjusted model comparing ≥6 months to never breastfeeding, demonstrating the added benefit of longer breastfeeding duration.

We found that longer breastfeeding duration was associated with reduced risk in biomarkers reflecting metabolic, cardiovascular, immune, and neuroendocrine systems. Specifically, longer breastfeeding duration was associated with significantly lower BMI, systolic and diastolic blood pressure, pulse, steeper cortisol slopes at T2 and T3, higher HDL, lower HbA1c at T2, and lower HS-CRP at T3. While our previous study only examined five of the same biomarkers, breastfeeding was similarly associated with better BMI, systolic blood pressure, and HDL, but in contrast was not significantly associated with diastolic blood pressure and HbA1c.¹⁶ The added detection of associations with diastolic blood pressure and HbA1c in the current study is likely due to having a more precise measure of breastfeeding and a larger proportion of Black and Hispanic participants in the sample, which increased the sensitivity of detecting the effect.

Prior research on metabolic syndrome and diabetes have similarly provided evidence of breastfeeding's beneficial influence on blood pressure, glucose metabolism, and body weight.^15,30,31 The beneficial effect of breastfeeding duration on maternal postpartum allostatic load, including its effect on specific biomarkers, provide evidence of pathways involving metabolic, cardiovascular, immune, and neuroendocrine systems that may contribute to reducing maternal chronic disease risk, including metabolic syndrome,¹⁵ diabetes,^14,32 fatty liver disease,³³ and breast and ovarian cancers.^12,34

In this study, although the interaction between breastfeeding and race/ethnicity was not statistically significant, race/ethnicity-stratified models showed that when comparing mothers who breastfed ≥6 months to those who never breastfed, there were greater reductions in allostatic load for Hispanic mothers at T2 and White mothers at T2 and T3, compared with Black mothers, for which inverse associations were not statistically significant. A previous analysis of CCHN data by Shalowitz et al¹⁰ found that while clinical measures of allostatic load were 28% and 31% lower at T2 and T3, respectively, for White mothers who breastfed compared with White mothers who never breastfed, allostatic load was only 6% (T2) and 5% (T3) lower for Hispanic mothers and 0.3% (T2) and 11% (T3) lower for Black mothers who breastfed, compared with respective mothers who never breastfed. Our study similarly found that breastfeeding had a consistent inverse association with allostatic load among White mothers.

However, Shalowitz et al¹⁰ did not adjust for confounders or incorporate breastfeeding duration, which likely explains our finding that there was a significant and large reduction in allostatic load among Hispanic mothers at T2 who breastfed ≥6 months compared with Hispanic mothers who never breastfed. Our previous study also found lower and nonsignificant reductions in allostatic load among Black and Hispanic mothers who were breastfeeding compared with White mothers who were breastfeeding.¹⁶

Current evidence indicating that breastfeeding has a favorable influence on allostatic load seems to be most consistent among White mothers and least consistent among Black mothers. This study offers some possible explanations for the weaker associations among Black mothers, given that we do not believe that breastfeeding is less protective of the health of Black mothers compared with White mothers and recognize race as a social construct with racial differences in allostatic load reflecting the imprinting of differential exposures to social and environmental stressors.³ First, prepregnancy allostatic load was not adjusted for in this study and in any previous analyses. Black women have been found to have higher allostatic load than White women at all ages²⁵ and consequently they likely entered pregnancy with a higher allostatic load. As such, the weaker inverse association may be explained by unmeasured higher baseline allostatic load as opposed to a lower moderating effect of breastfeeding on the allostatic load of Black mothers.

Second, the impact of intergenerational and persistent exposure to stressors such as discrimination, socioeconomic and neighborhood disadvantage, and other forms of systemic oppression that are salient in the lives of Black women,^35–38 may be so deep for Black mothers that unraveling the negative imprints on health requires greater effort and consistent exposure to protective factors. Third, breastfeeding may amplify psychosocial stressors for some women who breastfeed, particularly for mothers who live, work, and interact in environments and communities where breastfeeding is not the norm or not accepted/supported. Qualitative studies of Black women's experiences of breastfeeding have documented challenges faced in navigating unsupportive health care and work environments, and discouragement from family and friends.^39–43 These experiences may be physiologically taxing and may hinder mothers' capacities to fully reap the physiological benefits of breastfeeding.

Fourth, the biomarkers used to determine allostatic load may not carry the same validity in capturing health/risk status across different racial/ethnic groups due to the hegemony of patriarchy and Whiteness in their historical developments. A recent editorial in the Journal of American Medical Association reminded researchers to use caution in interpreting health-based risk scores such as the Framingham Risk Score due to the lack of inclusion of research participants from underrepresented groups in their development.⁴⁴ Despite these issues, this study observed a pattern of the protective influence of breastfeeding on allostatic load, which is evident across racial/ethnic groups although varying in strength. Based on study findings, future research on breastfeeding and allostatic load is strongly recommended to examine race/ethnicity-stratified models, even when interaction by race/ethnicity is not significant, so that social and environmental etiologies of health inequities can be better interrogated.

Strengths and limitations

An important limitation of this study was the inability to adjust for prepregnancy or baseline allostatic load, which was not collected in this CCHN sample. This prevented us from ruling out the possibility of reverse causation and may have led to an over- or underestimation of the association between breastfeeding and allostatic load. However, incorporation of prepregnancy BMI helped to refine our regression models. The data on breastfeeding did not capture duration beyond 6 months, intensity (i.e., exclusivity of breastfeeding versus combined formula and breastfeeding), and mode (i.e., feeding at the breast or bottle), which limited our ability to examine breastfeeding's influence more precisely, and may have led to an underestimation of the true effect of breastfeeding on allostatic load. Given that mothers reported breastfeeding at two time points and within the postpartum period, recall and information bias were likely minimized.

Response bias is a possibility with self-reported measures but was likely minimized by the rigorous questionnaire development process involving community input and the application of validated measures of stress exposure. Posthoc sensitivity analysis detected another limitation, which was the very small effect size (Cohen's f² = 0.002) observed among the Hispanic subgroup at 12 months postpartum indicating <80% power to examine associations in that model. However, estimated effect sizes from other models indicated sufficient power. An important strength of the study is the use of CBPR as part of the data collection process, which helped refine the questionnaires to appropriately capture community contexts and facilitated trust between researchers and participants so that data could accurately reflect participant experiences. Another strength of this study was the relatively large sample size of mothers from five regions across the United States and the relative high proportion of Black mothers, thus providing diversity and sufficient power to investigate the research questions.

Conclusion

This study determined that extended breastfeeding is a possible moderator of allostatic load, but further research is needed to elicit the association and mechanism of this relationship. Efforts are needed to recruit women prepregnancy and follow them postpartum to better analyze changes in allostatic load. Data on breastfeeding intensity and exclusivity over time are also important to better capture the influence of breastfeeding. This study builds on emergent literature on breastfeeding's impact on chronic stress, with an aim of promoting an inclusive understanding of protective factors for maternal health, particularly for Black women in the United States who experience greater health inequities.

Footnotes

Acknowledgments

The authors thank the Center for Research on Families and the Institute for Social Science Research for methodology consultation support.

Authors' Contributions

P.O. participated in the development of research tools and data collection protocol. B.H. analyzed data and wrote the article. B.H., H.L., L.S.C., P.O., and L.S. conceptualized the study and revised the article for critical intellectual content. All authors have read and approved the final content, and agreed to be accountable for all aspects of the work.

Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Supplementary Material

References

Ramey

, Schafer

, DeClerque

, et al. The preconception stress and resiliency pathways model: A multi-level framework on maternal, paternal, and child health disparities derived by community-based participatory research. Matern Child Health J, 2015; 19(4):707–719; doi: 10.1007/s10995-014-1581-1

O'Campo

, Schetter

, Guardino

, et al. Explaining racial and ethnic inequalities in postpartum allostatic load: Results from a multisite study of low to middle income women. SSM Popul Health, 2016; 2:850–858; doi: 10.1016/j.ssmph.2016.10.014

, Halfon

Racial and ethnic disparities in birth outcomes: A life-course perspective. Matern Child Health J, 2003; 7(1):13–30; doi: 10.1023/a:1022537516969

Riggan

, Gilbert

, Allyse

MA.

Acknowledging and addressing allostatic load in pregnancy care. J Racial Ethn Health Disparities, 2021; 8(1):69–79; doi: 10.1007/s40615-020-00757-z

McEwen

BS.

Stress, adaptation, and disease. Allostasis and allostatic load. Ann N Y Acad Sci, 1998; 840:33–44; doi: 10.1111/j.1749-6632.1998.tb09546.x

Beckie

TM.

A systematic review of allostatic load, health, and health disparities. Biol Res Nurs, 2012; 14(4):311–346; doi: 10.1177/1099800412455688

Guidi

, Lucente

, Sonino

, et al. Allostatic load and its impact on health: A systematic review. Psychother Psychosom, 2021; 90(1):11–27; doi: 10.1159/000510696

Johnson

, Cavallaro

, Leon

DA.

A systematic review of allostatic load in relation to socioeconomic position: Poor fidelity and major inconsistencies in biomarkers employed. Soc Sci Med, 2017; 192:66–73; doi: 10.1016/j.socscimed.2017.09.025

Hux

, Catov

, Roberts

JM.

Allostatic load in women with a history of low birth weight infants: The National Health and Nutrition Examination Survey. J Womens Health 2002, 2014; 23(12):1039–1045; doi: 10.1089/jwh.2013.4572

10.

Shalowitz

, Schetter

, Hillemeier

, et al. Cardiovascular and metabolic risk in women in the first year postpartum: Allostatic load as a function of race, ethnicity and poverty status. Am J Perinatol, 2019; 36(10):1079; doi: 10.1055/s-0038-1675618

11.

Gust

, Caccese

, Larosa

, et al. Neuroendocrine effects of lactation and hormone-gene-environment interactions. Mol Neurobiol, 2020; 57(4):2074–2084; doi: 10.1007/s12035-019-01855-8

12.

Victora

, Bahl

, Barros

AJD

, et al. Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. Lancet, 2016; 387(10017):475–490; doi: 10.1016/S0140-6736(15)01024-7

13.

Rameez

, Sadana

, Kaur

, et al. Association of maternal lactation with diabetes and hypertension: A systematic review and meta-analysis. JAMA Netw Open, 2019; 2(10):e1913401–e1913401; doi: 10.1001/jamanetworkopen.2019.13401

14.

Gunderson

, Lewis

, Lin

, et al. Lactation duration and progression to diabetes in women across the childbearing years: The 30-year CARDIA study. JAMA Intern Med, 2018; 178(3):328–337; doi: 10.1001/jamainternmed.2017.7978

15.

Ram

, Bobby

, Hailpern

, et al. Duration of lactation is associated with lower prevalence of the metabolic syndrome in midlife—SWAN, the study of women's health across the nation. Am J Obstet Gynecol, 2008; 198(3):268.e1–268.e6; doi: 10.1016/j.ajog.2007.11.044

16.

Hsiao

, Sibeko

Breastfeeding is inversely associated with allostatic load in postpartum women: Cross-sectional data from nationally representative US women. J Nutr, 2021; 151(12):3801–3810; doi: 10.1093/jn/nxab302

17.

Dunkel Schetter

, Schafer

, Lanzi

, et al. Shedding light on the mechanisms underlying health disparities through community participatory methods: The stress pathway. Perspect Psychol Sci, 2013; 8(6):613–633; doi: 10.1177/1745691613506016

18.

Morrison

, Shenassa

, Mendola

, et al. Allostatic load may not be associated with chronic stress in pregnant women, NHANES 1999–2006. Ann Epidemiol, 2013; 23(5):294–297; doi: 10.1016/j.annepidem.2013.03.006

19.

CDC. Results: Breastfeeding Rates. 2020. Available from: https://www.cdc.gov/breastfeeding/data/nis_data/results.html [Last accessed: September 30, 2020].

20.

CDC.. 2020 Breastfeeding Report Card. 2021. Available from: https://www.cdc.gov/breastfeeding/data/reportcard.htm [Last accessed: February 6, 2022].

21.

StataCorp. Stata Statistical Software: Release 17. 2021. College Station, TX: StataCorp LLC.

22.

Jones

, Kogan

, Singh

, et al. Factors associated with exclusive breastfeeding in the United States. Pediatrics, 2011; 128(6):1117–1125; doi: 10.1542/peds.2011-0841

23.

Odar Stough

, Khalsa

, Nabors

, et al. Predictors of exclusive breastfeeding for 6 months in a national sample of US children. Am J Health Promot, 2019; 33(1):48–56; doi: 10.1177/0890117118774208

24.

Crimmins

, Johnston

, Hayward

, et al. Age differences in allostatic load: An index of physiological dysregulation. Exp Gerontol, 2003; 38(7):731–734; doi: 10.1016/S0531-5565(03)00099-8

25.

Geronimus

, Hicken

, Keene

, et al. “Weathering” and age patterns of allostatic load scores among Blacks and Whites in the United States. Am J Public Health, 2006; 96(5):826–833; doi: 10.2105/AJPH.2004.060749

26.

Dunkler

, Plischke

, Leffondré

, et al. Augmented backward elimination: A pragmatic and purposeful way to develop statistical models. PLoS One, 2014; 9(11):e113677; doi: 10.1371/journal.pone.0113677

27.

Faul

, Erdfelder

, Buchner

, et al. Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behav Res Methods, 2009; 41(4):1149–1160; doi: 10.3758/BRM.41.4.1149

28.

Cohen

Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Routledge: New York, NY, USA; 2013.

29.

, Stuart

, Allison

DB.

Multiple imputation: A flexible tool for handling missing data. JAMA, 2015; 314(18):1966–1967; doi: 10.1001/jama.2015.15281

30.

Gunderson

, Lewis

, Wei

, et al. Lactation and changes in maternal metabolic risk factors. Obstet Gynecol, 2007; 109(3):729; doi: 10.1097/01.AOG.0000252831.06695.03

31.

Gunderson

, Hedderson

, Chiang

, et al. Lactation intensity and postpartum maternal glucose tolerance and insulin resistance in women with recent GDM: The SWIFT cohort. Diabetes Care, 2012; 35(1):50–56; doi: 10.2337/dc11-1409

32.

Aune

, Norat

, Romundstad

, et al. Breastfeeding and the maternal risk of type 2 diabetes: A systematic review and dose–response meta-analysis of cohort studies. Nutr Metab Cardiovasc Dis, 2014; 24(2):107–115; doi: 10.1016/j.numecd.2013.10.028

33.

Ajmera

, Terrault

, VanWagner

, et al. Longer lactation duration is associated with decreased prevalence of non-alcoholic fatty liver disease in women. J Hepatol, 2019; 70(1):126–132; doi: 10.1016/j.jhep.2018.09.013

34.

Chowdhury

, Sinha

, Sankar

, et al. Breastfeeding and maternal health outcomes: A systematic review and meta-analysis. Acta Paediatr, 2015; 104(S467):96–113; doi: 10.1111/apa.13102

35.

Williams

, Yu

, Jackson

, et al. Racial differences in physical and mental health: Socio-economic status, stress and discrimination. J Health Psychol, 1997; 2(3):335–351; doi: 10.1177/135910539700200305

36.

Williams

, Collins

Racial residential segregation: A fundamental cause of racial disparities in health. Public Health Rep 1974, 2001; 116(5):404–416; doi: 10.1093/phr/116.5.404

37.

Bartholomew

, Harris

, Maglalang

DD.

A call to healing: Black lives matter movement as a framework for addressing the health and wellness of black women. Community Psychol Global Perspect, 2018; 4(2):85–100; doi: 10.1285/i24212113v4i2p85

38.

Krieger

Methods for the scientific study of discrimination and health: An ecosocial approach. Am J Public Health, 2012; 102(5):936–944; doi: 10.2105/AJPH.2011.300544

39.

Spencer

, Wambach

, Domain

EW.

African American women's breastfeeding experiences: Cultural, personal, and political voices. Qual Health Res, 2015; 25(7):974–987; doi: 10.1177/1049732314554097

40.

Spencer

, Grassley

JS.

African American women and breastfeeding: An integrative literature review. Health Care Women Int, 2013; 34(7):607–625; doi: 10.1080/07399332.2012.684813

41.

Johnson

, Kirk

, Muzik

Overcoming workplace barriers: A focus group study exploring African American mothers' needs for workplace breastfeeding support. J Hum Lact, 2015; 31(3):425–433; doi: 10.1177/0890334415573001

42.

Johnson

, Kirk

, Rosenblum

, et al. Enhancing breastfeeding rates among African American women: A systematic review of current psychosocial interventions. Breastfeed Med, 2015; 10(1):45–62; doi: 10.1089/bfm.2014.0023

43.

Lutenbacher

, Karp

, Moore

ER.

Reflections of Black women who choose to breastfeed: Influences, challenges and supports. Matern Child Health J, 2016; 20(2):231–239; doi: 10.1007/s10995-015-1822-y

44.

Flanagin

, Frey

, Christiansen

, et al. Updated guidance on the reporting of race and ethnicity in medical and science journals. JAMA, 2021; 326(7):621–627; doi: 10.1001/jama.2021.13304

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.02 MB