Prenatal and Postnatal Experiences Predict Breastfeeding Patterns in the WIC Infant and Toddler Feeding Practices Study-2

Abstract

Objective:

To examine pre- and postnatal experiential factors associated with desirable breastfeeding patterns in a nationally representative population of low-income women who prenatally enrolled in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) and initiated breastfeeding.

Materials and Methods:

Using data from the longitudinal WIC Infant and Toddler Feeding Practices Study-2, multivariable, hierarchical logistic regression analyses identified prenatal and postnatal experiential factors associated with three breastfeeding patterns: (1) breastfeeding at 6 months, (2) breastfeeding at 1 year, and (3) breastfeeding at 1 year without introducing formula through age 6 months.

Results:

After controlling for covariates, one prenatal factor, breastfeeding intentions, and one postnatal factor, receipt of a doctor's recommendation to breastfeed, raised the odds of exhibiting the patterns analyzed. Another postnatal factor, returning to full-time employment before infant age 3 months, lowered the odds of exhibiting the patterns. Prior WIC participation significantly increased the odds of breastfeeding at 1 year, while postnatal employment before infant age 3 months significantly decreased the odds of exhibiting this pattern.

Conclusions:

Health care providers and those working in public health programs, including WIC, play an important role in helping low-income women mitigate shorter breastfeeding durations. Their efforts should continue focusing on bolstering women's prenatal breastfeeding intentions, reducing structural barriers to breastfeeding in the early postnatal period, particularly among those women returning to work, and connecting low-income families with WIC if they are not already enrolled in the program. This study is registered at clinicaltrials.gov as Feeding My Baby—A National WIC Study, NCT02031978.

Introduction

When examining breastfeeding duration, low-income women are at risk for shorter durations.^1,2 As such, there is a need to understand both pre- and postnatal experiential factors associated with breastfeeding patterns in this vulnerable population, especially the extent to which health care providers and public health interventions can help low-income women achieve breastfeeding durations that align with recommendations.

The Academy of Breastfeeding Medicine recommends that mothers exclusively breastfeed their infants for the first 6 months, followed by continued breastfeeding through at least 1 year along with feeding age-appropriate complementary foods.³ This recommendation is similar to those of other health organizations.^4–6 Yet, many women in the United States do not know the recommendations and fail to meet them.⁷ As a result, the U.S. Department of Health and Human Services has targeted the improvement of breastfeeding rates in several of its Healthy People (HP) objectives.

HP 2030 identifies two breastfeeding-related objectives. The first is to increase the proportion of infants who are exclusively breastfed through age 6 months, and the second is to increase the proportion of infants who are breastfed at age 1 year.⁸ The 2030 objectives indicate a continued need for community breastfeeding support. Additionally, given documented disparities in breastfeeding durations between lower and higher-income subpopulations,¹ a prior HP 2020 objective remains relevant: to increase the proportion of infants who are breastfed at age 6 months.

When examining factors associated with breastfeeding patterns, maternal experiences in the pre- and postnatal periods may be important. For example, receipt of a health care provider's support may promote longer durations among women who experience breastfeeding difficulties.⁹ Though pediatricians' endorsement of breastfeeding has increased substantially,¹⁰ the influence of their infant feeding recommendations on low-income women's breastfeeding durations remains unclear.

In addition to receiving a pediatrician's recommendation to breastfeed, participation in community interventions may also influence low-income women's breastfeeding durations. The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), a Federal public health nutrition program accessed by over 1.8 million low-income women¹¹ at nutritional risk, offers breastfeeding support among the benefits of program participation. WIC supports breastfeeding through breastfeeding education, peer counseling, and access to trained specialists to address breastfeeding challenges.¹² Despite these supports, WIC participants remain at high risk for short breastfeeding durations.¹

Breastfeeding outcomes in the WIC community have been a topic of research for years. Prior literature has reported that race, ethnicity, maternal age, prenatal infant feeding intentions, social support, and postpartum employment are associated with breastfeeding outcomes.^13–18 However, the nonrepresentative nature of the samples used in prior studies constrains the national relevance of many findings. For example, Gregory et al.¹⁸ used data from the Infant Feeding Practices Study II (IFPS II), fielded in 2005–2007, and found no association between WIC participation and breastfeeding at 3 months when prenatal feeding intentions were included in the analysis. While discussing their findings, the authors noted that the characteristics of WIC participants in the IFPS II sample differed notably from the national WIC population. Consequently, it is not clear whether findings using IFPS II data, or other subsets of the WIC population, accurately reflect factors associated with breastfeeding patterns in the larger, national group.

Given WIC's national reach and the national scope of the HP objectives, research utilizing a nationally representative sample is imperative. To our knowledge, only two studies have employed such data sets. The first, the WIC Infant Feeding Practices Study (WIC IFPS-1),¹⁹ was conducted in the mid-1990s and followed caregivers and their children (i.e., mother-child dyads) through the first year of life. The second, the WIC Infant and Toddler Feeding Practices Study-2 (WIC ITFPS-2), began in 2013²⁰ and is ongoing, following mother-child dyads until each child turns 9 years old, regardless of how long the dyad participates in WIC. Early WIC ITFPS-2 findings indicate that breastfeeding initiation rates and duration have markedly improved since WIC IFPS-1,^21,22 but durations still fall short of HP recommendations. Consequently, it is important to understand which experiential factors are associated with desirable breastfeeding patterns in the national WIC population, as these may drive programmatic efforts to boost breastfeeding in this at-risk population.

Using WIC ITFPS-2 data, this study examines the extent to which pre- and postnatal experiences are independently associated with three breastfeeding patterns that align with recent HP goals: breastfeeding at 6 months, breastfeeding at 1 year, and breastfeeding at 1 year without introducing infant formula in the first 6 months.

Materials and Methods

This study is secondary analysis of data originally collected as part of the national WIC ITFPS-2, a longitudinal cohort study following mother-child dyads from pregnancy or shortly after birth, through the child's ninth birthday. Though the study has been extended several times since the article was published, the original protocol is comprehensively described by Harrison et al.²⁰ In brief, WIC ITFPS-2 recruited participants from 80 WIC sites across 27 WIC State Agencies. Sites and their respective State Agencies were selected in a two-stage sampling process described by May et al.²¹ Sites were eligible for the study if they were expected to enroll at least 30 new study-eligible participants based on 2010 WIC Participant and Program Characteristics data.²³

WIC ITFPS-2 participants were recruited as they enrolled in the WIC program between July and November of 2013 at their respective WIC sites. To be eligible for WIC ITFPS-2, mothers had to be at least 16 years old, speak English or Spanish, and be enrolling in WIC for the first time for their pregnancy or for their infant no more than 2.5 months old. Consent to participate in WIC ITFPS-2 was obtained in person during enrollment in the study. The Westat Institutional Review Board (IRB) approved the national study under expedited authority. State and local IRBs approved local study activities as required by local policy.

This research leveraged the longitudinal design of WIC ITFPS-2 by using a nationally representative subsample of WIC ITFPS-2 participants: those who completed all study interviews between the prenatal period and 13-months postpartum (unweighted n = 1,459). Given our interest in postpartum employment's effect on breastfeeding behavior, we further excluded mothers who were less than 18 years old (i.e., minors) at the time of their child's birth, yielding a slightly smaller sample (unweighted n = 1,426). For the multivariate analyses, the subsample included adult, prenatal WIC enrollees who reported initiating breastfeeding (unweighted n = 1,207). Based on results from a power analysis, this sample size (unweighted n = 1,207) was sufficient to detect a small association between breastfeeding patterns analyzed and each of the covariates used in the logistic regression models with at least 90% power.

Data for this research were collected during eight telephone interviews. The first prenatal interview was completed in July 2013 and the last was completed in February of 2014. Because women can enroll in WIC at any time during pregnancy, interviewees varied in gestational weeks. The postnatal interviews were completed at 1-, 3-, 5-, 7-, 9-, 11-, and 13-months postpartum, with the last 13-month interview completed in August 2015.

Independent variables in the regression models included both sociodemographic and experiential factors. Sociodemographic variables included maternal race and ethnicity, maternal formal education level, maternal age at the child's birth, father's residence in the home, and receipt of Supplemental Nutrition Assistance Program (SNAP) benefits. Because self-report of income is known to contain a wide range of error,²⁴ receipt of SNAP, a means-tested Federal nutrition program, served as an indicator for relative incomes within the WIC population.

The focal experiential factors examined are categorized as occurring either pre- or postnatally. Prior WIC experience and intention to breastfeed are prenatal experiential factors. The Infant Feeding Intention (IFI) Scale assessed prenatal intentions to breastfeed.²⁵ Scores on the IFI Scale range from 0 to 16, with higher scores indicating stronger intention to breastfeed. Whether the infant's health care provider recommended breastfeeding and postpartum employment, either full-time or part-time, are postnatal experiences. The measure for employment reflects both the intensity of employment (part-time or full-time) and the postnatal timing, within 3 months postpartum or after 3 months postpartum.

Breastfeeding patterns were determined using responses to a single survey item at each postnatal interview: “Are you currently feeding breastmilk from the breast or a bottle, formula, or both?” Response options included: “only breastmilk,” “only formula,” and “both breastmilk and formula.” The response options allowed us to categorize respondents using the World Health Organization definition of breastfeeding.²⁶ According to this definition, infants are considered to be breastfed when they receive breastmilk (including milk expressed or from a wet nurse) and solid or semi-solid foods.

In the multivariate analyses, we modeled three different outcomes. Each aligns with HP objectives. Model 1 examines predictors of feeding only breastmilk or fed both breastmilk and formula until, at least, child age 6 months. This outcome aligns with the previous Maternal, Infant, and Child Health (MICH) HP 2020 objectives, which included increasing the proportion of infants who are breastfed at age 6 months.²⁷ We refer to this model as breastfeeding at 6 months.

Model 2 examines predictors of feeding only breastmilk or breastmilk and formula until, at least, child age 1 year. This outcome aligns with the HP 2030 objective MICH-16, increase the proportion of infants who are breastfed at 1 year.⁸ We refer to this model as breastfeeding at 1 year.

Model 3 examines predictors of feeding only breastmilk without introducing infant formula by child age 6 months and subsequently continuing to feed breastmilk until, at least, child 1 year. Given the phrasing of the survey item, the response “only breastmilk” indicates the absence of formula, not the absence of other complementary foods and beverages. We distinguish the third pattern from the one used in the second model because it makes use of the only response option consistent with “exclusive” breastfeeding; however, we do not use the term exclusive because the timing of the introduction of complementary foods and beverages is not known. While the third breastfeeding pattern does not directly align with a specific HP objective, study participants who exhibit this pattern may approximate the HP 2030 goal of exclusive breastfeeding for 6 months (MICH-15) and any breastfeeding through age 1 year (MICH-16).⁸ We refer to this model as breastfeeding at 1 year without introducing infant formula through age 6 months.

Data analysis

In all analyses, each case is weighted so that results reflect the national WIC population that met WIC ITFPS-2 eligibility criteria. The weighting appropriately accounts for sample selection and nonresponse bias. SAS statistical software package version 9.4 was used for all data manipulation and analyses.

Multivariable logistic regression models assessed the independent associations of participant characteristics with different breastfeeding patterns. To examine sources of variance, we took a hierarchical modeling approach, with prenatal and postnatal blocks of regressors added sequentially to allow for examination of the role of sociodemographic variables before introducing shared variance from experiential factors into the models. Statistical significance for all analyses was at the level of p < 0.05.

Results

Table 1 presents the sociodemographic characteristics of study participants used in the analysis. Tables 2 –4 report the odds ratios and 95% confidence intervals from the three multivariable regression models used to assess independent associations between participant characteristics and experiences, and specific breastfeeding patterns. In each table, Model A includes the same set of sociodemographic characteristics. Model B adds prenatal experiences to the set of sociodemographic characteristics. Model C is the final model, which includes sociodemographic characteristics, prenatal experiential factors, and postnatal experiential factors.

Table 1.

Frequency of Sociodemographic Characteristics and Experiential Factors Used in Regression Models Assessing Associations of Experiential Factors with Breastfeeding Patterns

Covariates	% (SE)	Unweighted n	Weighted n
Mother's ethnoracial group
Non-Hispanic Black	17.2 (2.6)	256	53,023
Non-Hispanic White	23.9 (3.2)	339	73,689
Non-Hispanic Other	6.6 (1.1)	70	20,400
Hispanic	52.3 (4.5)	542	161,030
Father's residence
Lives with the mother	60.6 (2.0)	709	186,711
Does not live with the mother	39.4 (2.0)	498	121,433
Mother's educational attainment
Less than high school	22.1 (2.4)	263	67,889
High school diploma or GED	36.0 (1.5)	421	110,420
Some college or vocational school	32.7 (2.2)	390	100,487
4-year college or some graduate work	9.2 (1.1)	129	28,176
Mother's intensity and postnatal timing of employment
Mother did not start work in first 13 months	38.3 (2.0)	453	118,002
Mother started part-time work after the first 3 months	19.1 (1.3)	224	58,837
Mother started full-time after the first 3 months	7.6 (1.0)	82	23,339
Mother started part-time work within the first 3 months	19.7 (1.5)	243	60,701
Mother started full-time work within the first 3 months	15.3 (1.3)	205	47,264
Infant's doctor's feeding recommendation
Recommended breastfeeding	45.6 (1.6)	540	139,968
Did not recommend breastfeeding	54.4 (1.6)	660	167,024
Mother's prior WIC experience
Has previously received benefits from WIC	55.4 (2.0)	677	170,482
Has not previously received benefits from WIC	44.6 (2.0)	529	137,460
SNAP participation
Receiving SNAP	44.3 (2.2)	551	136,521
Not receiving SNAP	55.7 (2.2)	655	171,508

GED, General Educational Development; SNAP, Supplemental Nutrition Assistance Program; WIC, Special Supplemental Nutrition Program for Women, Infants, and Children.

Table 2.

Odds Ratios and 95% Confidence Intervals from Hierarchical Logistic Regression of Breastfeeding at 6 Months on Sociodemographic and Experiential Factors (Model 1)

Independent variables	Model 1A: only sociodemographic characteristics OR (CI)^a	Model 1B: sociodemographic characteristic plus prenatal experiential factors OR (CI)^b	Model 1C:: sociodemographic characteristics plus pre- and postnatal experiential factors (final model) OR (CI)^c
Sociodemographics
Maternal ethnorace
Non-Hispanic White	Reference	Reference	Reference
Non-Hispanic Black	0.57 (0.36–0.89)	0.63 (0.41–0.97)	0.70 (0.44–1.12)
Non-Hispanic Other	1.17 (0.69–1.99)	1.04 (0.60–1.81)	1.03 (0.58–1.83)
Hispanic	0.93 (0.62–1.38)	0.90 (0.58–1.39)	0.86 (0.55–1.34)
Father's residence
Does not live with mother	Reference	Reference	Reference
Lives with mother	1.60 (1.18–2.17)	1.52 (1.06–2.16)	1.36 (0.96–1.93)
Educational attainment
Less than high school	Reference	Reference	Reference
High school diploma or GED	2.68 (1.46–4.94)	2.01 (1.03–3.93)	2.25 (1.09–4.67)
Some college or vocational school	1.10 (0.70–1.71)	0.85 (0.55–1.31)	0.94 (0.61–1.47)
4 Year college or some graduate work	1.19 (0.77–1.84)	1.06 (0.70–1.62)	1.10 (0.73–1.67)
SNAP^d
Not receiving SNAP	Reference	Reference	Reference
Receiving SNAP	0.77 (0.55–1.07)	0.75 (0.54–1.06)	0.66 (0.46–0.93)
Mother/caregiver's age (in years) when baby was born	1.05 (1.02–1.07)	1.05 (1.02–1.08)	1.05 (1.02–1.08)
Prenatal factors
WIC^e experience
Had not received benefits from WIC previously	Reference	Reference	Reference
Received benefits from WIC previously		1.14 (0.77–1.69)	1.14 (0.76–1.72)
IFI^f Scale score		1.25 (1.19–1.32)	1.24 (1.18–1.31)
Postnatal factors
Intensity and timing of employment
Not employed in the first 13 months			Reference
Part-time work after the first 3 months			0.92 (0.55–1.52)
Full-time after the first 3 months			0.50 (0.24–1.02)
Part-time work within the first 3 months			0.63 (0.41–0.96)
Full-time work within the first 3 months			0.37 (0.22–0.64)
Doctor's recommendation
Did not recommend breastfeeding			Reference
Recommended breastfeeding			1.54 (1.12–2.13)

Bold text represents statistically significant values.

Results are weighted. Weighted n = 305,395; unweighted n = 1,196. AIC = 379,140.

Results are weighted. Weighted n = 296,837; unweighted n = 1,162. AIC = 342,892.

Results are weighted. Weighted n = 295,891; unweighted n = 1,156. AIC = 333,163.

SNAP, Supplemental Nutrition Assistance Program.

WIC, Special Supplemental Nutrition Program for Women, Infants, and Children.

Prenatal IFI Scale sores range from 0 to 16, with higher scores indicating stronger intentions to breastfeed.

AIC, Akaike Information Criterion; CI, confidence interval; IFI, Infant Feeding Intentions; OR, odds ratio.

Table 3.

Odds Ratios and Confidence Intervals from Hierarchical Logistic Regression of Breastfeeding at 1 Year on Sociodemographic and Experiential Factors (Model 2)

Independent variables	Model 2A: only sociodemographic characteristics OR (CI)^a	Model 2B: sociodemographic characteristics plus prenatal experiential factors OR (CI)^b	Model 2C: sociodemographic characteristics plus pre- and postnatal experiential factors (final model) OR (CI)^c
Sociodemographics
Maternal ethnorace
Non-Hispanic White	Reference	Reference	Reference
Non-Hispanic Black	0.76 (0.45–1.30)	0.95 (0.56–1.61)	1.14 (0.61–2.13)
Non-Hispanic Other	1.56 (0.82–2.95)	1.60 (0.87–2.95)	1.66 (0.85–3.27)
Hispanic	1.18 (0.78–1.77)	1.25 (0.81–1.92)	1.27 (0.82–1.96)
Father's residence
Does not live with mother	Reference	Reference	Reference
Lives with mother	1.71 (1.18–2.48)	1.63 (1.11–2.39)	1.44 (0.98–2.12)
Educational attainment
Less than high school	Reference	Reference	Reference
High school diploma or GED	2.05 (1.24–3.37)	1.65 (1.03–2.65)	1.96 (1.14–3.36)
Some college or vocational school	1.15 (0.69–1.91)	0.92 (0.57–1.47)	1.06 (0.66–1.70)
4 Year college or some graduate work	1.29 (0.82–2.04)	1.22 (0.78–1.91)	1.31 (0.83–2.06)
SNAP^d
Not receiving SNAP	Reference	Reference	Reference
Receiving SNAP	0.77 (0.56–1.08)	0.71 (0.50–1.00)	0.59 (0.41–0.85)
Mother/caregiver's age (in years) when baby was born	1.05 (1.02––1.08)	1.05 (1.01–1.08)	1.05 (1.01–1.08)
Prenatal factors
WIC^e experience
Had not received benefits from WIC previously		Reference	Reference
Received benefits from WIC previously		1.38 (0.98–1.93)	1.43 (1.01–2.04)
IFI^f Scale score		1.28 (1.18–1.39)	1.27 (1.17–1.38)
Postnatal factors
Intensity and timing of employment
Not employed in the first 13 months			Reference
Part-time work after the first 3 months			1.00 (0.57–1.73)
Full-time after the first 3 months			0.33 (0.12–0.95)
Part-time work within the first 3 months			0.60 (0.37–0.97)
Full-time work within the first 3 months			0.30 (0.15–0.57)
Doctor's recommendation
Did not recommend breastfeeding			Reference
Recommended breastfeeding			1.71 (1.19–2.45)

Bold text represents statistically significant values.

Results are weighted. Weighted n = 305,395; unweighted n = 1,196. AIC = 324,880.

Results are weighted. Weighted n = 296,968; unweighted n = 1,162. AIC = 290,312.

Results are weighted. Weighted n = 295,891; unweighted n = 1,156. AIC = 278,077.

SNAP.

Special Supplemental Nutrition Program for WIC.

Prenatal IFI Scale sores range from 0 to 16, with higher scores indicating stronger intentions to breastfeed.

Table 4.

Odds Ratios and Confidence Intervals from Hierarchical Logistic Regression of Breastfeeding at 1 Year Without Formula Use in the First 6 Months on Sociodemographic and Experiential Factors (Model 3)

Independent variables	Model 3A: only sociodemographic characteristics OR (CI)^a	Model 3B: sociodemographic characteristics plus prenatal experiential factors OR (CI)^b	Model 3C: sociodemographic characteristics plus pre- and postnatal experiential factors (final model) OR (CI)^c
Sociodemographics
Maternal ethnorace
Non-Hispanic White	Reference	Reference	Reference
Non-Hispanic Black	0.35 (0.15–0.84)	0.43 (0.18–1.02)	0.52 (0.22–1.24)
Non-Hispanic Other	0.85 (0.41–1.76)	0.81 (0.38–1.76)	0.80 (0.31–2.05)
Hispanic	0.70 (0.46–1.07)	0.72 (0.46–1.13)	0.70 (0.43–1.13)
Father's residence
Does not live with mother	Reference	Reference	Reference
Lives with mother	1.91 (1.14–3.22)	1.75 (0.96–3.21)	1.52 (0.82–2.84)
Maternal educational attainment
Less than high school	Reference	Reference	Reference
High school diploma or GED	2.75 (1.21–6.24)	2.22 (0.91–5.41)	2.72 (1.07–6.89)
Some college or vocational school	1.72 (0.85–3.46)	1.44 (0.73–2.85)	1.64 (0.82–3.27)
4 Year college or some graduate work	1.57 (0.76–3.24)	1.60 (0.69–3.69)	1.68 (0.72–3.94)
SNAP^d receipt
Was not receiving SNAP	Reference	Reference	Reference
Receiving SNAP	0.80 (0.52–1.25)	0.74 (0.46–1.19)	0.62 (0.38–1.00)
Mother/Caregiver's age (in years) when baby was born	1.01 (0.97–1.04)	1.01 (0.96–1.05)	1.00 (0.96–1.05)
Prenatal factors
WIC^e experience
Had not received benefits from WIC previously		Reference	Reference
Received benefits from WIC previously		1.40 (0.86–2.27)	1.45 (0.88–2.39)
IFI^f Scale score		1.37 (1.26–1.49)	1.35 (1.23–1.49)
Postnatal factors
Intensity and timing of employment
Not employed in first 13 months			Reference
Part-time work after the first 3 months			1.03 (0.60–1.78)
Full-time after the first 3 months			0.60 (0.14–2.50)
Part-time work within the first 3 months			0.60 (0.31–1.14)
Full-time work within the first 3 months			0.18 (0.08–0.41)
Doctor's recommendation
Did not recommend reastfeeding			Reference
Recommended breastfeeding			1.88 (1.06–3.33)

Bold text represents statistically significant values.

Results are weighted. Weighted n = 305,095; unweighted n = 1,196. AIC = 243,201.

Results are weighted. Weighted n = 296,968; unweighted n = 1,162. AIC = 210,004.

Results are weighted. Weighted n = 296,022; unweighted n = 1,156. AIC = 200,016.

SNAP.

Special Supplemental Nutrition Program for WIC.

Prenatal IFI Scale scores range from 0 to 16, with higher scores indicating stronger intentions to breastfeed.

Model 1: Breastfeeding at 6 months (Table 2)

Thirty-seven percent of participants who initiated breastfeeding continued breastfeeding for at least 6 months. When only sociodemographic characteristics were examined (Model 1A), participants who self-identified as non-Hispanic Black were less likely than those who self-identified as non-Hispanic White to breastfeed 6 months. The father living with the mother as opposed to not living with the mother, having a high school education compared to having less than a high school education, and being older when the child was born increased the odds of breastfeeding 6 months. When the prenatal factors—prior WIC experience and IFI Scale scores—were added (Model 1B), these sociodemographic characteristics remained significant predictors. However, of the two prenatal experiential variables, only IFI Scale scores were significantly associated with this 6-month pattern, raising the odds of the outcome. When postnatal factors—level of work and a doctor's recommendation—were added (Model 1C), they were predictive. Postnatal employment occurring within 3 months postpartum, regardless of intensity, reduced the odds of this breastfeeding outcome compared to those not employed in the first 13 months of the child's life, while a recommendation to breastfeed from the infant's doctor increased the odds compared to those who did not receive this recommendation. In this final model, only three sociodemographic characteristics were significant: maternal education, maternal age, and SNAP receipt, with the latter indicating that those with lower incomes were less likely to be breastfeeding at 6 months than those with higher incomes.

Model 2: Breastfeeding at 1 year (Table 3)

Twenty-five percent of participants who initiated breastfeeding continued to breastfeed for at least 1 year. When only sociodemographic characteristics were examined (Model 2A), the father living in the home compared to not living in the home, a high school education compared to less than a high school education, and maternal age at the child's birth increased the odds of this 1-year outcome. When prenatal factors were included (Model 2B), IFI Scale scores were predictive, as were the previously mentioned sociodemographic characteristics. When postnatal experiential factors were included in the model (Model 2C), both were predictive. Full-time work, regardless of timing, reduced the likelihood of breastfeeding at 1 year compared to not working in the first 13 months of the child's life, and receipt of a doctor's recommendation increased the likelihood of this pattern compared to those who did not receive a recommendation to breastfeed. In this final model, prior WIC exposure increased the likelihood of the outcome compared to those who did not have prior experience.

Model 3: Breastfeeding at 1 year without introducing infant formula through age 6 months (Table 4)

Fifteen percent of participants who initiated breastfeeding indicated that they fed breastmilk without introducing infant formula for, at least, 6 months and continued to breastfeed until the child was, at least, 1 year of age. When only sociodemographic factors were examined (Model 3A), participants who self-identified as non-Hispanic Black were less likely to exhibit this pattern than participants who self-identified as non-Hispanic White. In addition, those who had the father residing in the home were more likely to exhibit this pattern than those who did not, as were those who had a high school education compared to those who had less than a high school education and older mothers. The proxy for income variation within the WIC community, SNAP receipt, was also predictive: those who received SNAP were less likely than those who did not to exhibit this pattern. Of the two prenatal experiential factors added to the model (Model 3B), only IFI Scale scores were positively associated with this breastfeeding outcome. When postnatal factors were added to the model (Model 3C), working full-time within 3 months postpartum reduced the likelihood of realizing this breastfeeding pattern compared to those who did not work, while a doctor's recommendation to breastfeed the infant increased the likelihood and IFI Scale scores remained positively associated with the outcome. Among the sociodemographic characteristics included in the final model, only maternal educational attainment was predictive.

Discussion

This study explored associations between four experiential factors and three desirable breastfeeding patterns among low-income women who enrolled in WIC prenatally and initiated breastfeeding. Among the prenatal influences explored, breastfeeding intentions were positively associated with each of the breastfeeding patterns, and prior WIC experience was positively associated with breastfeeding for 12 months. Among the postnatal factors explored, both having a doctor's recommendation to breastfeed and postpartum work status were significantly associated with each of the patterns, but in opposite directions. A doctor's recommendation to breastfeed consistently increased the odds of exhibiting the patterns analyzed compared to those who did receive this recommendation, while early postpartum employment consistently reduced the odds of exhibiting any of the patterns analyzed when compared to those not employed in the first 13 months of the child's life.

While the positive association between breastfeeding intentions and desirable breastfeeding patterns may not seem surprising, intentions do not always translate to behavior, especially in vulnerable populations.²⁸ Our findings indicate that among prenatal WIC enrollees strengthening breastfeeding intentions may improve the odds of achieving desirable breastfeeding patterns. These findings underscore the importance of comprehensive prenatal education that includes information on breastfeeding. Because intentions are strongly influenced by knowledge and support,²⁹ the WIC program should continue to invest in comprehensive prenatal breastfeeding education programs, including those such as breastfeeding peer counseling and the WIC Breastfeeding Support: Learn Together, Grow Together campaign. Likewise, physicians and other prenatal clinicians should ensure all low-income pregnant mothers receive access to high-quality, comprehensive breastfeeding education to maximize prenatal breastfeeding intentions.

This research also highlights the important role that pediatricians have in promoting longer breastfeeding durations among low-income women. Pediatrician encouragement of breastfeeding improved the odds of achieving each of the three outcomes analyzed. While research has indicated that some pediatricians are still hesitant to address the topic of infant feeding during the first month of the child's life,³⁰ a period that is pivotal for many breastfeeding mothers, it remains critical that pediatricians continue supporting and talking to low-income women about their breastfeeding decisions very early in the child's life.

The current study's findings also provide evidence suggesting that prior WIC program exposure plays a role in achieving longer breastfeeding durations among prenatal enrollees, after accounting for employment and input from the infant's doctor. While previous literature has found that prenatal WIC enrollees have better breastfeeding outcomes than postnatal enrollees,³¹ this study exclusively examines prenatal enrollees to better understand their specific experiences, and stratifies those women by prior WIC exposure while controlling for prenatal intentions. In this context, prior WIC exposure implicitly reflects at least two sources of accumulated experiential knowledge: knowledge gained from caring for at least one other child and knowledge gained from WIC participation. Both provide meaningful information for feeding subsequent children. Because these two factors are closely intertwined, they were not independently assessed in the models presented, which is a limitation of this study.

The finding that those with prior WIC program participation had higher odds of breastfeeding at 1 year than those without prior participation suggests that the benefits of WIC participation can have hold-over effects on breastfeeding subsequent children. In the current study, prior WIC exposure and prenatal feeding intentions were not significantly associated in bivariate analyses, suggesting that prior WIC exposure influences this outcome through pathways other than altering feeding intentions. For example, WIC provides breastfeeding support that may have a cumulative effect as women continue to learn and refine their personal strategies for achieving longer breastfeeding duration. As such, future research should investigate the extent to which the timing between WIC exposures influences breastfeeding outcomes, and should use qualitative methods to understand from the women's perspectives how repeated experiences with WIC may influence breastfeeding outcomes.

Early postpartum employment adversely affected breastfeeding duration; however, the degree to which postpartum employment influences breastfeeding depended on the intensity (part-time versus full-time) and timing (before versus after 3 months postpartum) of employment. Further work is needed to understand how WIC program supports (e.g., peer counseling, breastfeeding, and breast pump education) and workplace accommodations affect breastfeeding outcomes among WIC participants who become employed within the first year of the child's life.

The hierarchical regression approach taken in this study proved revealing when examining Model 3, the model with the outcome most closely aligned with the Academy's breastfeeding recommendation—that mothers exclusively breastfeed their infants for the first 6 months, followed by continued breastfeeding through at least 1 year along with age-appropriate complementary foods. Both ethnorace and father's residential status were initially significant predictors but became nonsignificant when prenatal experiential variables—specifically feeding intentions—were added to the model. When postnatal experiential variables were added, few sociodemographic characteristics remained predictive, but prenatal feeding intentions and postnatal experiential factors were significant. This unfolding pattern suggests that potentially modifiable experiences, both prenatal and postnatal, may ultimately be more important for this pattern of breastfeeding than sociodemographic characteristics.

In the 6-month model, ethnoracial characteristics became nonsignificant predictors when postnatal experiential factors were introduced into the model. This suggests that the postnatal experiential factors considered—the intensity and timing of postnatal employment and the infant's doctor's recommendation to breastfeed—share variance with ethnoracial characteristics and may mask disparities. Additional research is needed to examine whether characteristics of employment, including the nature of jobs and the breastfeeding supports offered by employers, differ by ethnoracial characteristics and whether the frequency or content of health care provider infant feeding recommendations differ by these characteristics. Furthermore, it is important to acknowledge race and ethnicity are in part social constructs, and the ethnoracial disparities observed in Models 1A and 1B are likely artifacts of longstanding systemic problems, including racial injustice and bias. To effectively address these disparities, structural barriers that disproportionately impact women of color must be removed, and more support must be appropriately directed to ensure equity is achieved.

This article focused on women who initiated breastfeeding because the vast majority, 84%, of this nationally representative group initiated breastfeeding. The minority who did not initiate breastfeeding may be a self-select group that may face unique challenges. Additional qualitative evidence is needed to adequately inform research on their motivations and constraints.

Additionally, while this study had many strengths, there are several limitations to the findings. First, though the sample weights address survey nonresponse bias, there may be participant characteristics for which they do not fully account. Second, the multivariate analyses considered many of the known factors associated with breastfeeding duration; however, other unmeasured variables might influence the outcomes of interest. Third, this study was unable to determine whether or how long participants exclusively breastfed their infants.

Conclusion

The findings presented demonstrate that health care providers and public health programs, such as WIC, have important roles in mitigating short breastfeeding durations among low-income women. The positive association between prior WIC experience and breastfeeding for at least 1 year suggests that repeated messaging through higher program dose may be important for achieving longer durations among low-income women. Given the findings, health care providers are encouraged to recommend breastfeeding to all low-income women, and to refer low-income families to WIC.

Authorship

C.B., N.S.W., and C.P. developed study question. C.B. oversaw metric development and data analysis. C.B. wrote the first drafts of the article with contributions from N.S.W. and C.P. All authors reviewed and commented on subsequent drafts of the article.

Ethical Standards Disclosure

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved 17 IRBs, including: Westat; state Department of Health IRBs in CA, CT, FL, GA, LA, MD, MI, NY, OH, OK, PA, SC, TN, and TX; and local IRBs at Arrowhead Regional Medical Center in San Bernardino, CA, and Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, CA. Use of these data does not imply that the IRBs, State Departments of Health, or WIC State Agencies and sites agree or disagree with any presentations, analyses, interpretations, or conclusions in this report. Written or verbal informed consent was obtained from all study participants. Verbal consent was witnessed and recorded.

Footnotes

Acknowledgments

The authors would like to thank Danielle Berman and Kelley Scanlon for providing guidance and feedback on the article.

Disclosure Statement

No competing financial interests exist.

Funding Information

This research was funded by the Office of Policy Support in the Food and Nutrition Service, USDA (contract AG-3198-B-11-0020). The findings and conclusions in this publication are those of the author(s) and should not be construed to represent any official USDA or U.S. Government determination or policy.

References

Centers for Disease Control and Prevention. Rates of any and exclusive breastfeeding by socio-demographics among children born in 2016. National Immunization Suvey. Available at https://www.cdc.gov/breastfeeding/data/nis_data/rates-any-exclusive-bf-socio-dem-2017.html (accessed April 6, 2021).

Milligan

, Pugh

, Bronner

, et al. Breastfeeding duration among low income women. J Midwifery Women Health, 2000; 45:246–252.

Chantry

, Eglash

, Labbok

. ABM position on breastfeeding—Revised 2015. Breastfeed Med, 2015; 10:407–411.

American Academy of

Pediatrics

, Section on

Breastfeeding

. Breastfeeding and the use of human milk. Pediatrics, 2012; 129:e827–e841.

Croke

. AAFP releases position paper on breastfeeding. Am Fam Physician, 2015; 91:56–57.

Lessen

, Kavanagh

. Position of the academy of nutrition and dietetics: Promoting and supporting breastfeeding. J Acad Nutr Diet, 2015; 115:444–449.

Wallenborn

, Ihongbe

, Rozario

, et al. Knowledge of breastfeeding recommendations and breastfeeding duration: A survival analysis on infant feeding practices II. Breastfeed Med, 2017; 12:156–162.

Office of Disease Prevention and Health Promotion. Healthy People 2020: Maternal, Infant, and Child Health. Available at https://health.gov/healthypeople/objectives-and-data/browse-objectives/infants (accessed April 10, 2021).

Gianni

, Bettinelli

, Manfra

, et al. Breastfeeding difficulties and risk for early breastfeeding cessation. Nutrients, 2019; 11:2266.

10.

Feldman-Winter

, Szucs

, Milano

, et al. National trends in pediatricians' practices and attitudes about breastfeeding: 1995 to 2014. Pediatrics, 2017; 140:e20171229.

11.

Kline

, Thorn

, Bellows

, et al. WIC participant and program characteristics 2018: U.S. Department of Agriculture, Food and Nutrition Service. 2020. Available at https://www.fns.usda.gov/wic/participant-and-program-characteristics-2018-charts (accessed April 17, 2021).

12.

U.S. Department of Agriculture ERS. WIC Program. August 20, 2019; Available at https://www.ers.usda.gov/topics/food-nutrition-assistance/wic-program/ (accessed April 16, 2021).

13.

Hurley

, Black

, Papas

, et al. Variation in breastfeeding behaviours, perceptions, and experiences by race/ethnicity among a low-income statewide sample of Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) participants in the United States. Matern Child Nutr, 2008; 4:95–105.

14.

Langellier

, Chaparro

, Whaley

. Social and institutional factors that affect breastfeeding duration among WIC participants in Los Angeles County, California. Matern Child Health J, 2012; 16:1887–1895.

15.

Schindler-Ruwisch

, Roess

, Robert

, et al. Determinants of breastfeeding initiation and duration among African American DC WIC recipients: Perspectives of recent mothers. Womens Health Issues, 2019; 29:513–521.

16.

Sparks

. Racial/ethnic differences in breastfeeding duration among WIC-eligible families. Womens Health Issues, 2011; 21:374–382.

17.

Tenfelde

, Finnegan

, Miller

, et al. Risk of breastfeeding cessation among low-income women, infants, and children: A discrete time survival analysis. Nurs Res, 2012; 61:86–95.

18.

Gregory

, Gross

, Nguyen

, et al. WIC participation and breastfeeding at 3 months postpartum. Matern Child Health J, 2016; 20:1735–1744.

19.

Baydar

, McCann

, Williams

, et al. WIC Infant Feeding Practices Study. Available at https://www.fns.usda.gov/wic/wic-infant-feeding-practices-study (accessed April 17, 2021).

20.

Harrison

, Hirschman

, Owens

, et al. WIC Infant and Toddler Feeding Practices Study: Protocol design and implementation. Am J Clin Nutr, 2014; 99:742S–746S.

21.

May

, Borger

, Weinfield

, et al. WIC Infant and Toddler Feeding Practices Study–2: Infant Year Report. Alexandria, VA: USDA Food and Nutrition Service, 2017. Available at https://www.fns.usda.gov/wic/wic-infant-and-toddler-feeding-practices-study-2-infant-year-report (accessed April 11, 2021).

22.

Paolicelli

, Berman

, Owens

, et al. Breastfeeding and complementary feeding patterns among women enrolled in WIC: WIC Infant and Toddler Feeding Practices Study-2. Breastfeed Med, 2017; 12:482–486.

23.

Connor

, Bartlett

, Mendelson

, et al. WIC Participant and Program Characteristics 2010, WIC-1-PC. US Department of Agriculture, Food and Nutrition Service, Alexandria, VA, 2011.

24.

Moore

, Stinson

, Welniak

. Income measurement error in surveys: A review. J Off Stat Stockholm, 2000; 16:331–362.

25.

Nommsen-Rivers

, Dewey

. Development and validation of the infant feeding intentions scale. Matern Child Health J, 2009; 13:334–342.

26.

World Health Organization. Indicators for assessing infant and young child feeding practices: Part 1: Definitions: Conclusions of a consensus meeting held 6–8 November 2007 in Washington DC, USA. World Health Organization, 2008.

27.

Office of Disease Prevention and Health Promotion. Maternal, infant, and child health. Available at https://www.healthypeople.gov/node/3492/data-details (accessed April 12, 2021).

28.

Kay

, Cholera

, Flower

, et al. Are low-income, diverse mothers able to meet breastfeeding intentions after 2 months of breastfeeding?. Breastfeed Med, 2020; 15:435–442.

29.

Mitra

, Khoury

, Hinton

, et al. Predictors of breastfeeding intention among low-income women. Matern Child Health J, 2004; 8:65–70.

30.

Meek

JY.

Pediatrician competency in breastfeeding support has room for improvement. Pediatrics 2017;140:e20172509 .

31.

Metallinos-Katsaras

, Brown

, Colchamiro

. Maternal WIC participation improves breastfeeding rates: A statewide analysis of WIC participants. Matern Child Health J, 2015; 19:136–143.