Abstract
Introduction: The health benefits of breastfeeding for both the mother and her offspring are well established, and breastfeeding may be especially important for mitigating negative health effects of pregnancy complicated by diabetes.
Objective: This study sought to examine the association between race/ethnicity, maternal diabetes, and breastfeeding initiation in South Carolina (SC).
Materials and Methods: Our study population is comprised of all live, singleton, full-term births in SC delivered January 2004 to December 2016 (292,468 non-Hispanic Whites, 173,158 non-Hispanic Blacks, and 52,174 Hispanics). Generalized estimating equations and an interaction term between race/ethnicity and maternal diabetes status were used to estimate the race/ethnic-specific odds ratio (OR) and 95% confidence intervals (CIs) that a mother with diabetes versus without diabetes initiated breastfeeding. Models were adjusted for maternal and infant sociodemographics (Model 1), additionally for clinical variables and birth outcomes (Model 2), and also for maternal prepregnancy body mass index (BMI) (Model 3).
Results: We found statistically significant differences in breastfeeding initiation by race/ethnicity and diabetes status in Model 1. These associations were attenuated and lost significance upon additional adjustment. Non-Hispanic Black mothers with gestational diabetes were significantly more likely to initiate breastfeeding than nondiabetic non-Hispanic Black mothers, even after adjustment for prepregnancy BMI (OR: 1.07 [95% CI 1.02–1.12]).
Conclusion: This large, population-based study of all live, singleton, full-term births in SC improves our understanding of how race/ethnicity and maternal metabolic disorders impact breastfeeding initiation, and may inform future hospital-based breastfeeding interventions in populations with the most need.
Keywords: gestational diabetes, maternal obesity, pregnancy, race/ethnic disparities
Introduction
Breastfeeding has been shown in many previous studies to confer tremendous health benefits upon both children and their mothers. Most recommendations conclude that mothers should breastfeed exclusively for ∼6 months, with continued breastfeeding for ∼1 year along with the introduction of appropriate complementary foods.1
When compared with infant formula, mature breast milk has a unique composition, including a higher fat and lower protein content as well as bioactive and immunogenic factors, and the act of breastfeeding as opposed to bottle feeding has been shown to reduce adverse offspring health outcomes.2 Lack of breastfeeding has been associated with increased infant respiratory and ear infections, adiposity, Type 2 diabetes, asthma, and dermatitis among many other adverse health outcomes for both offspring and mother.3 However, despite the scientific consensus that breastfeeding has a beneficial impact on both the mother and the infant, estimates show that ∼83% of U.S. infants were ever breastfed in 2015, with <60% being breastfed through 6 months and only 36% being breastfed through 12 months.4
Race/ethnic disparities in breastfeeding rates in the United States persist across all levels of breastfeeding initiation, exclusivity, and duration. Currently, non-Hispanic Black Americans have the lowest rates of breastfeeding, with only 64% of non-Hispanic Black mothers initiating breastfeeding, as compared with 82% of non-Hispanic White and Hispanic infants.5 Even after adjustment for known socioeconomic and behavioral risk factors, the racial/ethnic disparity in breastfeeding prevalence persists, suggesting a nuanced relationship not captured by standard confounding factors.6,7 These disparities are especially distressing as simulation studies suggest that suboptimal breastfeeding may be associated with excess morbidity, mortality, and monetary burden among minorities as compared with non-Hispanic Whites, with a 2017 study estimating 2.2 times the number of excess child deaths occurring among non-Hispanic Blacks, and 1.5 times the number of excess child deaths occurring among Hispanics, as compared with non-Hispanic Whites.8
Infants born to diabetic mothers may be at an increased risk of short- and long-term negative health outcomes, and studies suggest that adequate breastfeeding may alleviate some of this risk.3,9–12 Furthermore, women who develop gestational diabetes during a pregnancy may subsequently be at an increased risk of developing type 2 diabetes and other cardiometabolic disorders; breastfeeding has been shown to alleviate some of this increased risk.13 Thus, encouraging diabetic mothers to breastfeed their offspring remains crucial, especially given the overall lower rate of breastfeeding among diabetic mothers.13–15
However, the relationship between maternal diabetes and breastfeeding may be complicated by race/ethnicity, and understanding the interplay of these factors is important to inform targeted counseling of expectant mothers. There are not many studies examining race/ethnic disparities, diabetes, and breastfeeding, but several previous studies have been conducted examining other maternal metabolic disorders (e.g., maternal obesity) in this context.13–17 These studies have reported conflicting results, with some suggesting no difference in the relationship between maternal metabolic disorders and breastfeeding initiation by race/ethnicity, and others suggesting no relationship or a reversed association among specific race/ethnic groups.13–17
Thus, we sought to examine odds of breastfeeding initiation by race/ethnicity and maternal diabetes status across all live, singleton, full-term births in South Carolina (SC) occurring 2004–2016—a diverse state with ∼27% African Americans and 6% Hispanics.18
Materials and Methods
Study design and data collection
To examine this aim, we conducted a retrospective cohort study by merging three population-based data sources from SC: birth certificates, Medicaid, and inpatient hospital discharge diagnoses. Our dataset includes all live, singleton births occurring from January 2004 to December 2016 among SC residents. For mothers with pregnancy data available in Medicaid and the inpatient hospital discharge diagnosis datasets, we linked all data sources for up to a year before the birth to allow us to capture additional cases of diabetes.
Figure 1 presents the flow chart defining our study population. Exclusions included preterm births (gestational age <37 weeks) and a gestational age >44 weeks, implausible or extreme prepregnancy body mass index (BMI) and gestational weight gain, birthweight inconsistent with gestational age, maternal race/ethnicity identified as “other,” maternal age <16 years and >45 years, and a prepregnancy BMI <18.5 kg/m2. Implausible or extreme gestational weight gain values were defined as >99.5% or <0.5% of values for that variable. We eliminated 14 mothers with a prepregnancy BMI >95 g/m2 due to implausible weights for that BMI. Birthweight inconsistent with gestational age was defined according to criteria from Alexander et al.19,20
FIG. 1.
Flowchart of study sample size.
The Institutional Review Board at the Medical University of South Carolina approved this study. No written or verbal informed consent was required of participants as this study was conducted using existing health record datasets.
Outcome variable assessment
The outcome variable for this research was breastfeeding initiation (i.e., answering “Yes” to the question, “Is the infant being breastfed?”) as recorded in the birth certificates dataset.
Exposure variable assessment
Our exposures of interest were race/ethnicity and maternal diabetes. Maternal race/ethnicity was categorized as non-Hispanic White, non-Hispanic Black, or Hispanic. Maternal diabetes was defined as the presence of either pregestational diabetes or gestational diabetes as recorded in the birth certificate, inpatient hospital discharge diagnoses, or Medicaid datasets. For the Medicaid and inpatient hospital discharge diagnoses datasets, the following International Classification of Diseases (ICD)-9 and ICD-10 codes were used to define pregestational diabetes: 250.0–250.9, O24.0, O24.1, O24.3, O24.8, E10, and E11. The following ICD-9 and ICD-10 codes were used to define gestational diabetes: 648.0, 648.8, O24.4, and O24.9.
The interaction effect of these two exposures gave us the following categories of exposures: non-Hispanic White mother with pregestational diabetes, non-Hispanic White mother with gestational diabetes, non-Hispanic White mother without diabetes, non-Hispanic Black mother with pregestational diabetes, non-Hispanic Black mother with gestational diabetes, non-Hispanic Black mother without diabetes, Hispanic mother with pregestational diabetes, Hispanic mother with gestational diabetes, and Hispanic mother without diabetes. We drew comparisons by maternal diabetes status within each race/ethnicity; mothers without diabetes were the reference group.
Covariate variable assessment
Additional study variables were obtained from the linked datasets, and included sociodemographic and clinical variables. Continuous study variables obtained from the birth certificates included the following: gestational weight gain (kg), number of previous live births, maternal age (years), gestational age at birth (weeks), year of birth (2004–2016), and infant birthweight (kg).
Categorical study covariates obtained from the birth certificates included the following: admission to the neonatal intensive care unit (NICU), maternal prepregnancy BMI categories, maternal smoking pre- and during pregnancy, health insurance, prenatal care quality assessed using the revised GINDEX, maternal education, maternal hypertension pre- and during pregnancy, route of delivery (cesarean or vaginal), Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) participation, and county of residence.
Maternal prepregnancy BMI was categorized as normal weight (18.5 kg/m2 ≤ BMI <25.0 kg/m2), overweight (25.0 kg/m2 ≤ BMI <30.0 kg/m2), and obese (BMI ≥30.0 kg/m2). Normal weight prepregnancy BMI category was the reference category in analyses. Gestational weight gain was categorized as inadequate, adequate, or excessive according to the Institute of Medicine (IOM) guidelines based on prepregnancy BMI category.21 Adequate gestational weight gain was the reference category in analyses. Birthweight-for-gestational-age percentiles were calculated based on estimates from Talge et al., and categorized as large-for-gestational-age (LGA defined as >90th birthweight-for-gestational-age percentile) or small-for-gestational-age (SGA defined as <10th birthweight-for-gestational-age percentile).22 Health insurance was dichotomized as private/managed or Medicaid/other. Private/managed health insurance was the reference category for analyses. The Revised GINDEX classifies pregnancies as receiving inadequate, intermediate, adequate, intensive, or no care based on timing of care initiation and number of visits.23 Adequate prenatal care through the revised GINDEX was the reference category used in analyses. Maternal education was classified as less than high school, high school or equivalent, some college, associates or bachelor's degree, and master's or doctoral degree. Having less than a high school degree was the reference category in analyses.
Statistical analyses
SAS System (version 9.3; SAS Institute, Cary, NC) was used to run all analyses. A p-value of 0.05 and corresponding 95% confidence intervals (CIs) were used to determine statistical significance. Mann–Whitney U and chi-square tests were used to examine unadjusted bivariate associations between maternal diabetes status and population characteristics.
Generalized estimating equations with a random effect at the county level were used to estimate the odds of a mother in SC initiating breastfeeding. To examine the association between race/ethnicity, diabetes, and breastfeeding initiation in SC, a two-way interaction between race/ethnicity and diabetes was included in this model. We ran three models estimating odds of our outcome by this exposure: (1) a model adjusted for basic sociodemographics (Model 1: maternal age, health insurance, education, smoking status, gestational weight gain [IOM categories], revised GINDEX, number of previous live births, WIC participation, and year of birth); (2) a model additionally adjusted for birth outcomes and clinical variables (Model 2: Model 1 + maternal hypertension, route of delivery, large-for-gestational age, small-for-gestational age, gestational age, admission to NICU); and (3) a model additionally adjusted for maternal prepregnancy BMI categories (Model 3: Model 2 + prepregnancy BMI categories). Model 3 allowed us to assess the independent effect of maternal diabetes status on breastfeeding initiation. Final regression models gave us estimates of the overall adjusted odds ratio (OR) and 95% CI estimates of breastfeeding initiation by race/ethnicity and maternal diabetes status in SC.
Results
After applying exclusions, 517,800 mother–child dyads were included in our analysis. Figure 2 presents overall unadjusted trends in breastfeeding initiation by maternal diabetes status among our cohort stratified by race/ethnicity. From 2004 to 2016, the percentage of SC mothers initiating breastfeeding increased overall and for all racial/ethnic groups and regardless of diabetes status. Hispanic mothers experienced the smallest increase in breastfeeding initiation, and non-Hispanic Black and non-Hispanic White mothers both experienced over a 50% increase in breastfeeding initiation over time. Across all time points, Hispanics had the highest proportion of mothers initiating breastfeeding, and non-Hispanic Black mothers had the lowest proportion of mothers initiating breastfeeding. The rates of breastfeeding initiation were so high for Hispanic mothers that even Hispanic mothers with pregestational diabetes initiated breastfeeding at higher rates than non-Hispanic Blacks with or without diabetes. Mothers without diabetes had the highest rates of breastfeeding among non-Hispanic White and Hispanic mothers, though mothers with gestational diabetes had very similar rates to nondiabetic mothers. Among non-Hispanic Black mothers, the rates of breastfeeding initiation were highest among mothers with gestational diabetes. Overall, it appeared that mothers with pregestational diabetes had the lowest rates of breastfeeding initiation across all races/ethnicities. The overall proportion of SC mothers in our sample who initiated breastfeeding in 2016 was 76.8%. Table 1 presents the unadjusted bivariate association between maternal and infant characteristics and maternal race/ethnicity. All maternal and infant characteristics differed significantly by maternal race/ethnicity.
FIG. 2.
Temporal trends in breastfeeding initiation by race/ethnicity and maternal diabetes for all live, singleton, full-term births in South Carolina from 2004 to 2016.
Table 1.
Maternal and Infant Characteristics by Maternal Race/Ethnicity for All Live, Singleton, Full-Term Births in South Carolina from 2004 to 2016
| Non-Hispanic White (N = 292,468) | Non-Hispanic Black (N = 173,158) | Hispanic (N = 52,174) | Total (N = 51,7800) | p-Value | |
|---|---|---|---|---|---|
| Maternal age (years) [mean ± SD (N)] | 27.3 ± 5.8 (N = 292,466) | 25.1 ± 5.6 (N = 173,158) | 26.7 ± 5.8 (N = 52,172) | 26.5 ± 5.8 (N = 517,796) | <0.0001a |
| Previous live births [mean ± SD (N)] | 0.9 ± 1.1 (N = 292,437) | 1.1 ± 1.2 (N = 173,134) | 1.3 ± 1.2 (N = 52,161) | 1.0 ± 1.2 (N = 517,732) | <0.0001a |
| Clinically estimated gestation (weeks) [mean ± SD (N)] | 38.9 ± 1.0 (N = 292,468) | 38.8 ± 1.0 (N = 173,158) | 39.0 ± 1.1 (N = 52,174) | 38.9 ± 1.0 (N = 517,800) | <0.0001a |
| Maternal diabetes | |||||
| No diabetes | 264,467 (90.4%) | 154,931 (89.5%) | 46,429 (89.0%) | 465,827 (90.0%) | <0.0001b |
| Gestational diabetes | 23,989 (8.2%) | 13,838 (8.0%) | 4,943 (9.5%) | 42,770 (8.3%) | |
| Pregestational diabetes | 4,012 (1.4%) | 4,389 (2.5%) | 802 (1.5%) | 9,203 (1.8%) | |
| Prepregnancy BMI category | |||||
| Normal weight | 145,252 (57.5%) | 56,390 (40.3%) | 22,549 (51.3%) | 224,191 (51.4%) | <0.0001b |
| Overweight | 73,570 (29.1%) | 46,286 (33.1%) | 17,069 (38.8%) | 136,925 (31.4%) | |
| Obese | 33,670 (13.3%) | 37,288 (26.6%) | 4,358 (9.9%) | 75,316 (17.3%) | |
| Gestational weight gain (IOM guidelines) | |||||
| Inadequate | 53,217 (21.1%) | 45,480 (32.6%) | 14,060 (32.2%) | 112,757 (25.9%) | <0.0001b |
| Adequate | 75,179 (29.9%) | 36,572 (26.3%) | 14,019 (32.1%) | 125,770 (28.9%) | |
| Excessive | 123,374 (49.0%) | 57,254 (41.1%) | 15,629 (35.8%) | 196,257 (45.1%) | |
| Maternal hypertension | |||||
| No | 272,128 (93.0%) | 158,475 (91.5%) | 50,122 (96.1%) | 480,725 (92.8%) | <0.0001b |
| Yes | 20,340 (7.0%) | 14,683 (8.5%) | 2,052 (3.9%) | 37,075 (7.2%) | |
| Tobacco use | |||||
| No | 231,122 (79.1%) | 155,159 (89.7%) | 50,161 (96.2%) | 436,442 (84.3%) | <0.0001b |
| Yes | 61,186 (20.9%) | 17,896 (10.3%) | 1,987 (3.8%) | 81,069 (15.7%) | |
| Maternal health insurance | |||||
| Private/self-pay | 149,971 (51.6%) | 38,992 (22.7%) | 25,222 (49.2%) | 214,185 (41.7%) | <0.0001b |
| Medicaid/other | 140,874 (48.4%) | 132,847 (77.3%) | 25,996 (50.8%) | 299,717 (58.3%) | |
| Maternal education | |||||
| <High school | 40,435 (13.8%) | 37,753 (21.9%) | 29,621 (57.1%) | 107,809 (20.9%) | <0.0001 |
| High school | 65,902 (22.6%) | 56,760 (32.9%) | 10,537 (20.3%) | 133,199 (25.8%) | |
| Some college | 70,085 (24.0%) | 50,805 (29.4%) | 5,987 (11.5%) | 126,877 (24.6%) | |
| College degree | 115,699 (39.6%) | 27,404 (15.9%) | 5,704 (11.0%) | 148,807 (28.8%) | |
| NICU admission | |||||
| No | 285,241 (97.5%) | 168,988 (97.6%) | 51,199 (98.1%) | 505,428 (97.6%) | <0.0001b |
| Yes | 7,227 (2.5%) | 4,170 (2.4%) | 975 (1.9%) | 12,372 (2.4%) | |
| Mode of delivery | |||||
| Vaginal | 197,217 (67.4%) | 115,980 (67.0%) | 37,865 (72.6%) | 351,062 (67.8%) | <0.0001b |
| Cesarean section | 95,251 (32.6%) | 57,178 (33.0%) | 14,309 (27.4%) | 166,738 (32.2%) | |
| Revised GINDEX | |||||
| Inadequate | 25,399 (8.7%) | 25,365 (14.8%) | 12,206 (23.6%) | 62,970 (12.2%) | <0.0001b |
| Adequate | 71,934 (24.7%) | 30,243 (17.6%) | 6,502 (12.6%) | 108,679 (21.1%) | |
| Intermediate | 141,973 (48.8%) | 82,714 (48.2%) | 27,928 (54.0%) | 252,615 (49.1%) | |
| Intensive | 51,584 (17.7%) | 33,348 (19.4%) | 5,075 (9.8%) | 90,007 (17.5%) | |
| Mother received WIC benefits | |||||
| No | 170,735 (59.2%) | 39,477 (23.2%) | 17,312 (33.9%) | 227,524 (44.6%) | <0.0001b |
| Yes | 117,694 (40.8%) | 131,031 (76.8%) | 33,804 (66.1%) | 282,529 (55.4%) | |
| Large-for-gestational-age | |||||
| No | 258,638 (88.4%) | 165,209 (95.4%) | 47,102 (90.3%) | 470,949 (91.0%) | <0.0001b |
| Yes | 33,830 (11.6%) | 7,949 (4.6%) | 5,072 (9.7%) | 46,851 (9.0%) | |
| Small-for-gestational-age | |||||
| No | 272,737 (93.3%) | 146,816 (84.8%) | 48,369 (92.7%) | 467,922 (90.4%) | <0.0001b |
| Yes | 19,731 (6.7%) | 26,342 (15.2%) | 3,805 (7.3%) | 49,878 (9.6%) | |
Mann–Whitney U test for significant differences by race/ethnicity.
Chi-square test for significant differences by race/ethnicity.
NICU, neonatal intensive care unit; WIC, Special Supplemental Nutrition Program for Women, Infants, and Children; IOM, Institute of Medicine; BMI, body mass index; SD, standard deviation.
The unadjusted intraclass correlation coefficient for the county-level random effect was 0.099, suggesting that 9.9% of the variability in breastfeeding initiation in this population was accounted for by county of residence. Table 2 presents the ORs and 95% CIs for the association between maternal and infant characteristics and breastfeeding initiation. All variables except for intermediate and intensive prenatal care, maternal overweight prepregnancy, and maternal diabetes were significantly associated with breastfeeding initiation across all models. Maternal age, education, excessive gestational weight gain, year of birth, gestational age, and offspring being large-for-gestational age at birth were positively associated with breastfeeding initiation. Tobacco use, Medicaid/other health insurance, inadequate gestational weight gain, inadequate prenatal care, previous live births, WIC participation, maternal hypertension, Cesarean Section delivery, NICU admission, and offspring being small-for-gestational age at birth were negatively associated with breastfeeding initiation.
Table 2.
Odds Ratios and 95% Confidence Intervals (OR [95% CI]) for Breastfeeding Initiation by Maternal and Infant Characteristics for All Live, Singleton, Full-Term Births in South Carolina from 2004 to 2016
| Characteristics | Model 1a | Model 2b | Model 3c |
|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Primary exposure | |||
| Non-Hispanic White | |||
| Gestational diabetes | 0.94 (0.90–0.97) | 0.97 (0.93–1.00) | 0.98 (0.94–1.01) |
| Pregestational diabetes | 0.85 (0.78–0.93) | 0.91 (0.84–0.99) | 0.93 (0.86–1.01) |
| Non-Hispanic Black | |||
| Gestational diabetes | 1.03 (0.98–1.08) | 1.05 (1.01–1.10) | 1.07 (1.02–1.12) |
| Pregestational diabetes | 0.93 (0.86–1.00) | 0.99 (0.92–1.07) | 1.02 (0.94–1.10) |
| Hispanic | |||
| Gestational diabetes | 0.91 (0.82–1.00) | 0.93 (0.84–1.02) | 0.93 (0.85–1.03) |
| Pregestational diabetes | 0.76 (0.61–0.95) | 0.80 (0.64–1.01) | 0.81 (0.65–1.02) |
| Maternal and infant characteristics | |||
| Maternal age (years)d | 1.01 (1.01–1.01) | 1.01 (1.01–1.02) | 1.02 (1.01–1.02) |
| Medicaid/other health insurance | 0.77 (0.76–0.79) | 0.77 (0.76–0.79) | 0.77 (0.76–0.79) |
| Maternal high school education | 1.37 (1.34–1.40) | 1.36 (1.34–1.40) | 1.37 (1.34–1.40) |
| Maternal some college | 2.27 (2.22–2.32) | 2.26 (2.21–2.31) | 2.26 (2.21–2.31) |
| Maternal college degree | 3.68 (3.58–3.79) | 3.63 (3.53–3.74) | 3.63 (3.52–3.73) |
| Tobacco use | 0.64 (0.63–0.66) | 0.65 (0.64–0.67) | 0.65 (0.64–0.67) |
| Inadequate gestational weight gain | 0.88 (0.86–0.89) | 0.89 (0.87–0.90) | 0.89 (0.87–0.91) |
| Excessive gestational weight gain | 1.05 (1.03–1.07) | 1.04 (1.02–1.06) | 1.05 (1.03–1.07) |
| Inadequate prenatal care | 0.85 (0.83–0.87) | 0.83 (0.81–0.86) | 0.83 (0.81–0.86) |
| Intermediate prenatal care | 1.02 (1.00–1.04) | 1.00 (0.99–1.03) | 1.00 (0.98–1.02) |
| Intensive prenatal care | 1.00 (0.97–1.02) | 0.99 (0.96–1.01) | 0.99 (0.98–1.02) |
| Previous live birthsd | 0.86 (0.85–0.87) | 0.86 (0.85–0.86) | 0.86 (0.85–0.86) |
| Mother received WIC benefits | 0.79 (0.78–0.81) | 0.79 (0.78–0.81) | 0.80 (0.78–0.81) |
| Year of birthd | 1.06 (1.06–1.07) | 1.07 (1.06–1.07) | 1.07 (1.06–1.07) |
| Maternal hypertension | — | 0.91 (0.89–0.94) | 0.93 (0.90–0.96) |
| Cesarean section delivery | — | 0.88 (0.87–0.90) | 0.89 (0.88–0.91) |
| Large-for-gestational age | — | 1.08 (1.05–1.11) | 1.08 (1.05–1.11) |
| Small-for-gestational age | — | 0.89 (0.87–0.91) | 0.89 (0.87–0.91) |
| Gestational age (weeks)d | — | 1.06 (1.05–1.07) | 1.06 (1.05–1.07) |
| NICU admission | — | 0.82 (0.78–0.86) | 0.82 (0.78–0.86) |
| Maternal overweight prepregnancy | — | — | 1.02 (1.00–1.04) |
| Maternal obese prepregnancy | — | — | 0.93 (0.91–0.95) |
Bolded values represent p ≤ 0.05.
Model adjusted for maternal age, insurance, education, smoking status, gestational weight gain categories, prenatal care quality, number of previous live births, whether mom received WIC benefits, and year of birth.
Model adjusted for maternal age, insurance, education, smoking status, gestational weight gain categories, prenatal care quality, number of previous live births, whether mom received WIC benefits, year of birth, admission to the NICU, SGA, LGA, gestational age, and maternal hypertension.
Model adjusted for maternal age, insurance, education, smoking status, gestational weight gain categories, prenatal care quality, number of previous live births, whether mom received WIC benefits, year of birth, admission to the NICU, SGA, LGA, gestational age, maternal hypertension, and prepregnancy BMI categories.
ORs and 95% CIs represent a 1-unit increase.
OR, odds ratio; CI, confidence interval.
Among non-Hispanic Whites, maternal pregestational and gestational diabetes were associated with significantly lower odds of breastfeeding initiation as compared with nondiabetics. This association was attenuated and lost significance upon further adjustment. Among Hispanics, maternal pregestational diabetes was associated with significantly lower odds of breastfeeding initiation as compared with nondiabetics in Model 1, but no other significant associations were observed for this race/ethnicity. Among non-Hispanic Black mothers, maternal gestational diabetes was associated with significantly higher odds of breastfeeding initiation as compared with nondiabetics in Model 2 and Model 3, but not for models only adjusted for maternal and infant sociodemographics.
Discussion
In our population of live, singleton, full-term births in SC from 2004 to 2016, we observed an increasing trend in breastfeeding initiation, which mirrors that seen in other population-based U.S. studies.24,25 According to the U.S. Centers for Disease Control, 83.2% of infants born in 2015 were ever breastfed, up from 73.1% of infants born in 2004.26 Our study shows substantially lower rates of breastfeeding initiation across all time points, from 49.4% in 2004 to 76.8% in 2016. During this time frame, SC began statewide and local efforts to improve breastfeeding rates, including incentives for achieving Baby-Friendly USA® hospital designation and prioritization of breastfeeding by WIC services. Despite these initiatives, breastfeeding initiation among these live, singleton, full-term births in SC was still 5% points below the Healthy People 2020 goal of 81.9% in 2016.27
We report substantial differences in breastfeeding initiation by maternal race/ethnicity in our study population. Non-Hispanic Black mothers were the least likely to breastfeed, and Hispanic mothers were the most likely to breastfeed in our cohort. We found breastfeeding initiation rates similar to those reported at the national level, though we observed a slightly larger difference between breastfeeding initiation rates among non-Hispanic Whites and Hispanics.5 This large difference in breastfeeding initiation rates echoes those seen in other studies, suggesting that it is due to cultural and/or socioeconomic factors.6,7 In adjusted models, we report significantly lower odds of breastfeeding initiation among mothers with pregestational diabetes for non-Hispanic Whites and Hispanics. However, upon further adjustment for confounding this association was no longer significant. This suggests that factors related to clinical and birth outcomes (e.g., mode of delivery or infant gestational age), as well as maternal prepregnancy obesity, may be confounding associations observed between diabetes and breastfeeding initiation in prior studies.13–15,25,28,29
Although there have been several studies on the impact of maternal metabolic disorders on breastfeeding initiation, there is not a lot of cohesive evidence related to the impact of gestational diabetes status on breastfeeding initiation.13–15,25,28,29 In mothers with gestational diabetes, lactation may be more difficult due to some of the same mechanisms that potentially impact mothers with other metabolic disorders: delayed onset of lactogenesis, lower physiologic levels of prolactin in response to suckling, lower colostrum expression, a lower intention to breastfeed due to differences in breastfeeding efficacy and knowledge, requirement for higher risk postpartum care, and maternal discomfort breastfeeding in front of others.30–32
However, despite concerns about lowered breastfeeding initiation rates in this population, a recent population-based study of diabetes status and breastfeeding initiation using Ohio birth certificate data reported that mothers with gestational diabetes were more likely to initiate breastfeeding than both mothers with pregestational diabetes and nondiabetic mothers.13 This is similar to our study, which reported either no significant differences in breastfeeding initiation or significantly higher odds of breastfeeding initiation among mothers with gestational diabetes versus no diabetes. In both studies, an underlying factor driving this association may be the increased interaction mothers with gestational diabetes have with the medical system or differences in clinical counseling, including recommended additional counseling on the benefits of breastfeeding.13,33,34
However, mothers who received intensive or intermediate care according to the revised GINDEX—which takes into account timing of care initiation and number of visits—were not significantly more likely to initiate breastfeeding in any of our models. These results suggest that other factors may be driving this observed association. Additional research is needed to understand whether the lack of significant differences in breastfeeding initiation between mothers with gestational diabetes versus no diabetes may be due to hospital-level factors (e.g., Baby-Friendly initiatives), and how to ensure successful long-term, exclusive breastfeeding in these mothers.35,36
Importantly, when models in our study were mutually adjusted for prepregnancy BMI category and diabetes, most associations between diabetes status and breastfeeding initiation were nonsignificant. This finding suggests that maternal prepregnancy BMI category—specifically maternal obesity, which resulted in significantly reduced odds of breastfeeding initiation in our study—may be a more important factor impacting breastfeeding initiation in our population. Our results are unsurprising given the strong association between obesity and pregestational and gestational diabetes, but stand in contrast to those reported by Matias et al., who reported an independent effect of maternal prepregnancy BMI and gestational diabetes on breastfeeding initiation rates.37,38 After adjusting for prepregnancy overweight and obesity, the positive association between gestational diabetes status and odds of breastfeeding initiation among non-Hispanic Blacks was strengthened. This suggests that—independent of maternal prepregnancy obesity—non-Hispanic Black mothers with gestational diabetes are initiating breastfeeding at higher rates in SC. To understand the role insulin–glucose metabolism plays in breastfeeding initiation, future research should examine the interplay between maternal diabetes and prepregnancy BMI, and include information on diabetes treatment and control.
There are several limitations to this population-based retrospective cohort study. First, generalizability of the SC data outside of the context of the Southeastern United States is likely not possible, given its unique sociodemographic makeup and breastfeeding initiatives. Second, we excluded offspring who were not born full term, and included several variables in our models (e.g., large-for-gestational-age (LGA), small-for-gestational-age (SGA), gestational age, NICU admission) that may have been potential mediators of our association of interest, which may have biased our results toward the null. However, we still observed significant associations between our exposures and outcomes of interest, suggesting that differences in breastfeeding initiation by race/ethnicity and maternal diabetes status do not operate solely through these birth outcomes. Our study may also be limited by the quality of the birth certificate data, although the 2004 revision of the birth certificate improved data quality. Estimates for the accuracy of exposures and outcomes in the birth certificate vary widely, with a sensitivity as low as 46% in Ohio for gestational diabetes to 95% accuracy in Massachusetts for breastfeeding initiation.39,40 However, we were also able to ascertain maternal diabetes status from Medicaid and hospital discharge datasets, allowing us to capture additional cases. Another limitation to the birth certificate datasets includes their lack of information on duration or exclusivity of breastfeeding, which are important factors impacting offspring health.41 Some women in our study may have had multiple births during the study period. We treated these births in our analyses as independent events, which assume that breastfeeding behaviors do not remain the same for a mother across all pregnancies. This assumption may not be valid, though we noticed a strong significant association between lower odds of breastfeeding initiation with increasing live births, suggesting that breastfeeding initiation rates do not remain the same for a mother across all pregnancies. Finally, we did not have access to oral glucose tolerance test values or data reflecting adequacy of diabetic control, and thus cannot examine hyperglycemia as a continuum for our exposure. However, the diagnosis of diabetes is what informs clinical intervention, and thus we feel confident our study will be applicable within the context of SC's clinical and public health practice.
Conclusion
This retrospective cohort study makes use of three large population-based datasets and >10 years of data collection, making it one of the largest studies of this association. We report an overall breastfeeding initiation rate of 76.8% across all live, singleton, full-term births in SC in 2016, and increases in breastfeeding initiation occurring among all race/ethnic groups from 2004 to 2016. We report significant differences in breastfeeding initiation by race/ethnicity and diabetes status, and no independent effect of pregestational or gestational diabetes on breastfeeding initiation after adjustment for prepregnancy BMI category among non-Hispanic Whites and Hispanics. Among non-Hispanic Black mothers, individuals with gestational diabetes were significantly more likely to initiate breastfeeding than nondiabetics, even after adjustment for prepregnancy BMI category. This study improves our understanding of how race/ethnicity and maternal metabolic disorders impact breastfeeding initiation, and may inform future hospital-based breastfeeding interventions in populations with the most need.
Disclosure Statement
No competing financial interests exist.
Funding Information
This research has been funded by the Medical University of South Carolina's Clinical and Translational Research Institute (SCTR) through NIH Grant Number UL1 TR001450 and NIH-NCATS TL1 TR001451.
References
- 1. Victora CG, Bahl R, Barros AJD, et al. Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. Lancet 2016;387:475–490 [DOI] [PubMed] [Google Scholar]
- 2. Oddy W. Commentary infant feeding and obesity risk in the child. Breastfeed Rev 2012;20:7–12 [PubMed] [Google Scholar]
- 3. Ojha S, Robinson L, Symonds ME, et al. Suboptimal maternal nutrition affects offspring health in adult life. Early Hum Dev 2013;89:909–913 [DOI] [PubMed] [Google Scholar]
- 4. Centers for Disease Control and Prevention's (CDC) Division of Nutrition, Physical Activity, and Obesity National Center for Chronic Disease Prevention and Health Promotion. Breastfeeding report card United States. 2018. Available at www.cdc.gov/breastfeeding (accessed January13, 2019)
- 5. Anstey EH, Chen J, Elam-Evans LD, et al. Racial and geographic differences in breastfeeding—United States, 2011–2015. MMWR Morb Mortal Wkly Rep 2017;66:723–727 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Fryer K, Santos HP, Pedersen C, et al. The Hispanic paradox: Socioeconomic factors and race/ethnicity in breastfeeding outcomes. Breastfeed Med 2018;13:174–180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Singh GK, Kogan MD, Dee DL. Nativity/immigrant status, race/ethnicity, and socioeconomic determinants of breastfeeding initiation and duration in the United States, 2003. Pediatrics 2007;119(Suppl 1):S38–S46 [DOI] [PubMed] [Google Scholar]
- 8. Bartick MC, Jegier BJ, Green BD, et al. Disparities in breastfeeding: Impact on maternal and child health outcomes and costs. J Pediatr 2017;181:49–55 [DOI] [PubMed] [Google Scholar]
- 9. Godfrey K, Reynolds R, Prescott S, et al. The influence of maternal obesity on the long-term health of the offspring. Lancet Diabetes Endocrinol 2017;5:53–64 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Yu Z, Han S, Zhu J, et al. Pre-pregnancy body mass index in relation to infant birth weight and offspring overweight/obesity: A systematic review and meta-analysis. PLoS One 2013;8:e61627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kawasaki M, Arata N, Miyazaki C, et al. Obesity and abnormal glucose tolerance in offspring of diabetic mothers: A systematic review and meta-analysis. PLoS One 2018;13:1–19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Crume TL, Ogden L, Maligie M, et al. Long-term impact of neonatal breastfeeding on childhood adiposity and fat distribution among children exposed to diabetes in utero. Diabetes Care 2011;34:641–645 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Kachoria R, Oza-Frank R. Differences in breastfeeding initiation by maternal diabetes status and race, Ohio 2006–2011. Matern Child Health J 2014;18:2226–2235 [DOI] [PubMed] [Google Scholar]
- 14. Thompson LA, Zhang S, Black E, et al. The association of maternal pre-pregnancy body mass index with breastfeeding initiation. Matern Child Health J 2013;17:1842–1851 [DOI] [PubMed] [Google Scholar]
- 15. Masho S, Cha S, Morris M. Prepregnancy obesity and breastfeeding noninitiation in the United States: An examination of racial and ethnic differences. Breastfeed Med 2015;10:253–262 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Wojcicki J. Maternal prepregnancy body mass index and initiation and duration of breastfeeding: A review of the literature. J Womens Health 2011;20:341–347 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Kachoria R, Oza-Frank R. Factors associated with breastfeeding at discharge differ by maternal diabetes type. Diabet Med 2014:1222–1229 [DOI] [PubMed] [Google Scholar]
- 18. United States Census Bureau. QuickFacts. 2017. Available at https://census.gov/quickfacts/fact/table/US/PST045217 (accessed January22, 2019)
- 19. Parker JD, Schoendorf KC. Implications of cleaning gestational age data. Paediatr Perinat Epidemiol 2002;16:181–187 [DOI] [PubMed] [Google Scholar]
- 20. Alexander GR, Himes JH, Kaufman RB, et al. A United States national reference for fetal growth. Obstet Gynecol 1996;87:163–168 [DOI] [PubMed] [Google Scholar]
- 21. Hrolfsdottir L, Rytter D, Olsen SF, et al. Gestational weight gain in normal weight women and offspring cardio-metabolic risk factors at 20 years of age. Int J Obes 2015;39:671–676 [DOI] [PubMed] [Google Scholar]
- 22. Talge NM, Mudd LM, Sikorskii A, et al. United States birth weight reference corrected for implausible gestational age estimates. Pediatrics 2014;133:844–853 [DOI] [PubMed] [Google Scholar]
- 23. Alexander GR, Kotelchuck M. Quantifying the adequacy of prenatal care: A comparison of indices. Public Health Rep 111:408–418; discussion 419 [PMC free article] [PubMed] [Google Scholar]
- 24. Herrick KA, Rossen LM, Kit BK, et al. Trends in breastfeeding initiation and duration by birth weight among US children, 1999–2012. JAMA Pediatr 2016;170:805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kachoria R, Moreland JJ, Cordero L, et al. Trends in breastfeeding initiation, continuation, and exclusivity by maternal prepregnancy weight: 2004–2011. Obesity 2015;23:1895–1902 [DOI] [PubMed] [Google Scholar]
- 26. Centers for Disease Control and Prevention. Breastfeeding report card, United States. 2018. Available at https://cdc.gov/breastfeeding/data/reportcard.htm (accessed May22, 2019)
- 27. US Department of Health and Human Services. Healthy people 2020. Available at https://healthypeople.gov/2020/topics-objectives/topic/family-planning (accessed February2, 2019)
- 28. Cordero L, Thung S, Landon MB, et al. Breast-feeding initiation in women with pregestational diabetes mellitus. Clin Pediatr (Phila) 2014;53:18–25 [DOI] [PubMed] [Google Scholar]
- 29. Cordero L, Gabbe SG, Landon MB, et al. Breastfeeding initiation in women with gestational diabetes mellitus. J Neonatal Perinatal Med 2013;6:303–310 [DOI] [PubMed] [Google Scholar]
- 30. Nommsen-Rivers LA. Does insulin explain the relation between maternal obesity and poor lactation outcomes? An overview of the literature. Adv Nutr 2016;7:407–414 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Park S, Lee JL, In Sun J, et al. Knowledge and health beliefs about gestational diabetes and healthy pregnancy's breastfeeding intention. J Clin Nurs 2018;27:4058–4065 [DOI] [PubMed] [Google Scholar]
- 32. Doughty KN, Ronnenberg AG, Reeves KW, et al. Barriers to exclusive breastfeeding among women with gestational diabetes mellitus in the United States. J Obstet Gynecol Neonatal Nurs 2018;47:301–315 [DOI] [PubMed] [Google Scholar]
- 33. Loewenberg Weisband Y, Rausch J, Kachoria R, et al. Hospital supplementation differentially impacts the association between breastfeeding intention and duration among women with and without gestational diabetes mellitus history. Breastfeed Med 2017;12:338–344 [DOI] [PubMed] [Google Scholar]
- 34. Oza-Frank R, Gunderson EP. In-hospital breastfeeding experiences among women with gestational diabetes. Breastfeed Med 2017;12:261–268 [DOI] [PubMed] [Google Scholar]
- 35. Oza-Frank R, Moreland JJ, McNamara K, et al. Early lactation and infant feeding practices differ by maternal gestational diabetes history. J Hum Lact 2016;32:658–665 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Nguyen PTH, Binns CW, Nguyen CL, et al. Gestational diabetes mellitus reduces breastfeeding duration: A prospective cohort study. Breastfeed Med 2019;14:39–45 [DOI] [PubMed] [Google Scholar]
- 37. Pinheiro TV, Goldani MZ IV.APSA Group. Maternal pre-pregnancy overweight/obesity and gestational diabetes interaction on delayed breastfeeding initiation. PLoS One 2018;13:e0194879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Matias SL, Dewey KG, Quesenberry CP, et al. Maternal prepregnancy obesity and insulin treatment during pregnancy are independently associated with delayed lactogenesis in women with recent gestational diabetes mellitus. Am J Clin Nutr 2014;99:115–121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Bodnar LM, Abrams B, Siminerio L, et al. Validity of birth certificate-derived maternal weight data in twin pregnancies. Matern Child Nutr 2016;12:632–638 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Dietz P, Bombard J, Mulready-Ward C, et al. Validation of selected items on the 2003 U.S. standard certificate of live birth: New York city and Vermont. Public Health Rep 2015;130:60–70 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Spatz DL. Preventing obesity starts with breastfeeding. J Perinat Neonatal Nurs 2014;28:41–50 [DOI] [PubMed] [Google Scholar]