Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Nov 1.
Published in final edited form as: Health Psychol. 2020 Sep 24;39(11):997–1006. doi: 10.1037/hea0001028

Multilevel Prenatal Socioeconomic Determinants of Mexican American Children’s Weight: Mediation by Breastfeeding

Sarah G Curci 1, Juan C Hernández 1, Linda J Luecken 1, Marisol Perez 1
PMCID: PMC7919009  NIHMSID: NIHMS1673104  PMID: 32969698

Abstract

Objective:

Mexican American (MA) children are more likely to grow up in poverty than their non-Hispanic/Latinx white peers and are at an elevated risk for early onset obesity. The current study evaluated the effects of prenatal family- and neighborhood-level disadvantage on children’s weight and weight gain from 12 months through 4.5 years of age. Maternal breastfeeding duration was evaluated as a potential mechanism underlying the relation between multilevel disadvantage and weight.

Methods:

Data was collected from 322 low-income, MA mother-child dyads. Women reported the degree of family socioeconomic disadvantage and breastfeeding status. Neighborhood disadvantage was evaluated with census-level metrics. Children’s weight and height were measured at laboratory visits.

Results:

Greater prenatal neighborhood disadvantage predicted higher child Body Mass Index (BMI) at 12 months, over and above family-level disadvantage; this effect remained stable through 4.5 years. Breastfeeding duration partially mediated the effect of neighborhood disadvantage on child BMI. Breastfeeding duration predicted child BMI at all timepoints.

Conclusions:

Maternal prenatal residence in a neighborhood with high concentrated disadvantage may place low-income, MA children at increased risk of elevated weight status during the first few years of life. Breastfeeding duration emerged as potentially modifiable pathway through which the prenatal neighborhood impacts children’s early life weight.

Keywords: prenatal environment, neighborhood disadvantage, childhood obesity, breastfeeding

Childhood obesity is a well-documented risk factor for several maladaptive health outcomes in childhood and adulthood, including metabolic syndrome, diabetes, cardiovascular dysfunction, and cancer (Biro & Wien, 2010). Recent data suggest that nearly one in five children in the United States have obesity (Hales, Carroll, Fryar, & Ogden, 2017), yet there exist significant ethnic and racial disparities. The prevalence of obesity among Mexican American children (21.2%) is significantly higher than among non-Hispanic White children (14%; Ogden, Carroll, Kit, & Flegal, 2012; Singh, Siahpush, & Kogan, 2010). Disparities in overweight and obesity prevalence among low-income and minority children are evident as early as the toddler and preschool period (Anderson & Whitaker, 2009; Ogden et al., 2012). Early in life conditions, such as poverty and related stressors, elevate the risk of obesity and related health conditions (Conroy, Sandel, & Zuckerman, 2010; Chen & Miller, 2013). According to the U.S. Census Bureau’s 2011 American Community Survey, 28% of Hispanic children live in poverty at the family level, compared to 12% of non-Hispanic White children (Kids Count Data Center, 2017). Hispanic children are nearly five times more likely to live in a high poverty neighborhood than their non-Hispanic white peers (Kids Count Data Center, 2019). Socioeconomic inequalities in obesity rates continue to increase in the U.S., with low-income and minority children showing the highest obesity rates of all groups (Chung et al., 2016; Anderson & Butcher, 2006; Ogden et al., 2012). Awareness of the multiple risk exposures surrounding children developing in low-income families and neighborhoods has increased interest in identifying modifiable predictors for this population. A mechanistic understanding of poverty’s influence on early onset obesity, at both the family and neighborhood level, is essential for addressing health disparities among low-income, Mexican American children.

At the family level, low-income families are more likely to experience food insecurity, or the inability to attain nutritionally appropriate foods due to financial, social, or physical barriers (Anderson, 1990). Food insecurity can influence weight trajectories as caregivers choose larger quantitates of cheaper and poor quality (i.e., nutrient-sparse, energy-dense) foods over high quality (i.e., nutrient-dense, energy-sparse) more expensive foods (Hamilton et al., 1997; Bickel, Andrews, & Klein, 1996). At the neighborhood level, socioeconomic inequalities contribute to between-neighborhood differences in social and built environments, which impact obesity via more proximate factors (e.g., health behaviors; food markets; Schulz & Northridge 2004; Dubowitz et al., 2012; Inagami, Cohen, Brown, & Asch, 2009). A growing body of literature demonstrates several aspects of the socioeconomic, built, and social environment of a neighborhood contributes to unhealthy weight status during later childhood (Carroll-Scott et al., 2013; Fang, Thomsen, Nayga, & Goudie, 2018; Saelens et al., 2018), yet little is known about the effects of neighborhood disadvantage on children’s weight and potential proximate mediating factors during the earliest years of life.

Given the developing brain and body’s heightened susceptibility to influence from environmental exposures, the developmental origins of early onset disparities in obesity are likely rooted in processes beginning in the prenatal and postpartum period (Shonkoff & Phillips, 2000). A prenatal programming perspective suggests that environmental influences first begin in utero, initiating predispositions that set the stage for health disparities later in development (Nicholas et al., 2016). High prenatal stress has been linked to offspring overweight, obesity, and metabolic dysfunction in both animal and human research (Cao-Lei, Laplante, & King, 2016). The neighborhood context, specifically neighborhood concentrated disadvantage, may also be a relevant environmental context in the prenatal programming of physical health via maternal stress (King, Kane, Scarbrough, Hoyo, & Murphy, 2016). Early infancy and preschool have been implicated as important periods setting the foundation for lifelong physical health. For example, rapid weight gain in the first year of life is a predictor of subsequent obesity (Baird, et al., 2005). Existing research underscores the importance of considering the prenatal environment at both the family- and neighborhood-level (Glass & Bilal, 2016). Thus, a prospective, longitudinal approach beginning in the prenatal period is essential in discerning when and how family- and neighborhood-level disadvantage influence trajectories of early onset obesity.

Of particular interest in preventive efforts is the identification of modifiable factors underlying the relation between socioeconomic disadvantage and children’s obesity risk. One potential mechanism relevant to the perinatal and infancy periods may be breastfeeding. Among the well-known health benefits of breastfeeding, many studies provide evidence that breastfeeding is associated with a lower risk of obesity in childhood and later in life (Horta & Victora, 2013). Systematic literature reviews further suggest a direct relation between a longer duration of breastfeeding and a lower risk of later-life obesity, even after adjustment for potential confounds such as parental obesity (Owen, Martin, Whincup, Smith, & Cook, 2005). More recently, Guerrero et al. (2016) reported that breastfeeding was associated with a lower growth trajectory of BMI from four to six years of age. Singhal and Lanigan (2007) describe a “growth acceleration” theory in which a faster rate of growth during infancy adversely programs physiological parameters associated with metabolic syndrome, including obesity and insulin resistance. Breastfeeding, however, is hypothesized to slow early life acceleration of growth that lowers the risk of later life obesity and its associated health risks.

Given the numerous health benefits, the World Health Organization (WHO), Center for Disease Control (CDC), and American Academy of Pediatrics identify breastfeeding as a key strategy for promoting global health, with recommendations that women breastfeed exclusively for six months. However, in the United States, only approximately 42% of women breastfeed exclusively or in combination with formula-feeding through six months (CDC, 2010). Rates of breastfeeding show significant variability based on socioeconomic status and ethnicity. Approximately 45% of Hispanic mothers breastfeed to six months, compared to 27% of Black non-Hispanic mothers, while higher educated and higher income women are considerably more likely to breastfeed than WIC-eligible women or those with less than a high school education (CDC, 2010). At the family level, low-income mothers may be more likely to return to work sooner than their higher income peers and may be more likely to be employed in positions that make breastfeeding at work difficult (Shealy, Li, Benton-Davis, & Grummer-Strawn, 2005). Additionally, women with low educational attainment or who receive limited prenatal care may be less informed of the benefits and techniques of breastfeeding (CDC, 2011).

Societal and cultural barriers to breastfeeding, however, exist beyond the family level. Recent literature suggests that neighborhood disadvantage may influence the likelihood of a mother breastfeeding (Burdette, 2013). Disadvantaged neighborhoods may lack a variety of resources (e.g., support groups, safety, role models) related to initiating/maintaining breastfeeding (Silverman, Decker, Reed & Raj, 2006; Cunradi, Caetano, Clark, & Schafer, 2000; Kaufman, Deenadayalan, & Karpati, 2010). Conversely, living in a more advantaged neighborhood can serve to increase breastfeeding behavior, particularly among highly educated neighborhoods with higher rates of integrated maternal and infant health care and maternal social support (Alvarado, Zepeda, Rivero, Rico, López, & Díaz, 1999). Among low-income urban mothers, higher educated and more affluent neighborhood contexts are associated with more breastfeeding behaviors over and above individual social and financial resources (Leclair, Robert, Sprague, & Fleming, 2015; Daoud et al., 2014). Yourkavitch, Kane, and Miles (2018) found that the odds of any breastfeeding decreased as neighborhood disadvantage increased only among Hispanic women, suggesting that living in a disadvantaged neighborhood may be especially detrimental to breastfeeding practices of Hispanic women compared to other racial and ethnic groups of women.

Current Study

The current longitudinal study evaluated the impact of multilevel socioeconomic disadvantage during pregnancy on offspring BMI from 12 months to 54 months of age in a sample of impoverished Mexican American children. The within-group approach highlights the heterogeneity in weight status that exists among Mexican American children and elucidates potential mechanistic factors. The study takes a broad, contextual approach by evaluating disadvantage across two systemic levels: family and neighborhood. We evaluated breastfeeding duration as a potential modifiable mechanism through which family-level economic disadvantage and neighborhood-level concentrated disadvantage affect children’s weight and weight gain across the first few years of life. We hypothesized that higher family-level disadvantage and neighborhood-level disadvantage would be associated with higher child BMI at 12 months and increased weight gain through 54 months. Breastfeeding duration was expected to mediate the relation of disadvantage to children’s weight and weight gain, such that less family and neighborhood disadvantage would be associated with longer breastfeeding duration, and in turn, healthier weight and weight gain.

Method

Participants

The current sample includes 322 mother-child dyads participating in a larger, longitudinal study of maternal and child development in low-income Mexican-origin families in the southwestern United States. At the time of enrollment, women were between 18 and 42 years of age (M = 27.8) and had completed on average 10 years of education. The modal family income was $10,001 – $15,000 for an average household of four people. The majority of women were born in Mexico (86.0%), spoke Spanish as their primary language (82.0%), and were unemployed (83.5%). Most women were not first-time mothers (77.8%); of those women, the number of other biological children ranged from one to nine. See Table 1 for full sample demographics at enrollment.

Table 1.

Sample Demographics at Prenatal Visit

Maternal age; range, M (SD) 18–42; 27.8 (6.5)
Maternal country of birth; N (%)
 Mexico 278 (86%)
 United States 44 (14%)
Marital status; N (%)
 Married/living together 249 (77%)
 Single, never married 49 (15%)
 Separated/divorced 24 (7%)
Maternal education; N (%)
 Less than high school 190 (59%)
 High school diploma/GED 86 (27%)
 Some college/technical school/college degree 45 (14%)
Annual family income; N (%)
 <$5,000 44 (14%)
 $5,000–$10,000 61 (19%)
 $10,000–$15,000 87 (28%)
 $15,000–$20,000 37 (12%)
 $20,000–$25,000 40 (13%)
 $25,000 and above 45 (14%)
Missing/did not answer (n = 8)
Open in a new tab

Procedure

The study was approved by the Institutional Review Boards at Arizona State University and Maricopa Integrated Health System. Pregnant women were recruited from hospital-based prenatal clinics. Eligibility criteria included self-identification as Mexican or Mexican American, fluency in English or Spanish, at least 18 years of age, low-income status (self-reported family income below $25,000 or eligibility for Medicaid or Federal Emergency Services coverage for childbirth), and anticipated delivery of a singleton baby without prenatal evidence of health or developmental problems. Written informed consent from participants was obtained at the initial interview. Data for the current analyses were collected at one prenatal home visit (26–38 weeks gestation), five postpartum home visits (6 weeks, 12 weeks, 18 weeks, 24 weeks), and five laboratory visits (12 months, 18 months, 24 months, 36 months, and 54 months). Interviews were conducted in participants’ choice of Spanish (86%) or English (14%). Informed consent and survey questions were read aloud, and participants were provided visual aids with written and graphic descriptions of item responses. At the home visits, participants were compensated $50 and small gifts for the baby. Participants were compensated $100 and up to $50 for travel expenses for each laboratory visit.

The study employed a planned missingness design for home visits in the first six months after the birth in order to minimize participant burden and promote retention. The entire sample was assigned to complete the prenatal and 6-week visits, but participants were randomly assigned to miss either the 12-, 18-, or 24-week home visits. Of the 322 participants, 312 (97%) completed the 6-week visit, 203 (99% of those randomly assigned) completed the 12-week visit, 209 (96% of those randomly assigned) completed the 18-week visit, and 209 (93% of those randomly assigned) completed the 24-week visit. Two hundred and five participants (64%) completed the 12-month lab visit, 237 (73%) completed the 18-month lab visit, 243 (75%) completed the 24-month lab visit, 216 (67%) completed the 36-month lab visit, and 230 (71%) completed the 54-month lab visit. Participants with missing data at the 6-, 12-, 18-, and 24-week visits did not differ from those without missing data on study variables. Mothers born in the United States were more likely to have missing data at the 12-month lab visit, t(319) = 3.010, p = .003; 18-month lab visit, t(319) = 3.925, p < .001; 24-month lab visit, t(319) = 3.560, p < .001; 36-month lab visit, t(319) = 4.066, p < .001; and 54-month lab visit, t(319) = 3.057, p = .006.

Measures

Family economic disadvantage.

Women reported on their family’s economic stress at the prenatal home visit using the 20-item Economic Hardship Scale (EHS; Barrera, Caples, & Tein, 2001). The EHS is comprised of four subscales: financial strain, inability to make ends meet, not enough money, and economic adjustments. Scores on individual subscales are converted to z-scores and then summed to capture overall economic stress. Women rate how often and whether they expect to face or have faced various economic stressors. Higher scores indicate more economic stress. Internal consistency was acceptable (Cronbach’s alpha = .72).

Neighborhood concentrated disadvantage.

Census tract information for each participant was calculated using home addresses at the prenatal visit. Neighborhood data was pulled from the 2010 U.S. Census and 2012 American Community Survey (ACS; U.S. Census Bureau, 2011). Neighborhood concentrated disadvantage was calculated as a composite of five standardized, census-level metrics of neighborhood-level socioeconomic disadvantage: the percentage of residents living below the poverty line, the percentage of residents on public assistance, the percentage of female-headed households, the percentage of unemployed residents, and percentage of residents under the age of 18 years (Sampson, Raudenbush, & Earls, 1997). Internal consistency was acceptable (Cronbach’s alpha = .79).

Duration of breastfeeding.

Women reported on their breastfeeding status at the 6-, 12-, 18-, and 24-week home visits by responding to a single item “Do you breastfeed or bottle-feed your baby?” Women were given the following response options: breastfeed, bottle-feed, or both. Women who responded “both” were categorized as partial breastfeeding status. To capture women with any level of breastfeeding, breastfeeding status at each timepoint was coded 0 (bottle-feeding only) or 1 (partial or exclusive breastfeeding). Breastfeeding status at the four timepoints was then summed to calculate breastfeeding duration (0 = no breastfeeding, 1 = partial or full breastfeeding through 6 weeks, 2 = partial or full breastfeeding through 12 weeks, 3 = partial or full breastfeeding through 18 weeks, 4 = partial or full breastfeeding through 24 weeks). The calculation of a breastfeeding duration variable addresses the infant’s length of exposure to any breastfeeding, rather than the presence or absence at any single time point or the overall quantity of breastmilk received by the baby.

We chose to focus our analyses on breastfeeding status coded in this manner, rather than exclusive versus nonexclusive breastfeeding as suggested by Labbok and Krasovec (1990), because of its greater relevance to this group of mothers. Prior research has noted the common practice of combined breast- and bottle-feeding (“los dos”) among low-income, Mexican-origin mothers based on the belief that formula provides essential vitamins and promotes a healthier baby (Bunik et al., 2006). Indeed, at the prenatal visit only 6.8% of our sample reported an intent to exclusively breastfeed, and rates of actual exclusive breastfeeding were 13% at 6 weeks postpartum, 10.8% at 12 weeks, 15.5% at 18 weeks, and 12.1% at 24 weeks. However, we present additional analyses of breastfeeding duration with breastfeeding status at each timepoint coded as 0 = bottle-feeding or partial breastfeeding, or 1 = exclusive breastfeeding, and the time points summed as described above to calculate a duration variable.

Child body mass index (BMI).

Children’s weight and height were measured at the 12-, 18-, 24-, 36-, and 54-month lab visits and used to calculate child BMI (kg/m2). BMI-for-age percentile was calculated using an SPSS program provided by the World Health Organization using 2006 WHO tables and data (available at https://www.who.int/childgrowth/software/en/). BMI-for-age values for the current sample were compared to the WHO’s Child Growth Standards for children from birth to 5 years (WHO Multicentre Growth Reference Study Group, 2006). Raw score BMI was used in all analyses.

Covariates.

Maternal country of birth, child birth weight, and maternal BMI were included as potential covariates given documented relations with primary study variables. Among Mexican American families with U.S.-born mothers, children’s unhealthy weight status has been associated with higher maternal BMI (Hernández-Valero et al., 2007). Low birth weight has been associated with less maternal breastfeeding (Hediger, Overpeck, Ruan, & Troendle, 2000). Women reported their country of birth at the prenatal home visit (coded 1 = U.S., 2 = Mexico). Child birth weight was obtained through medical record review at the hospital of birth. Maternal weight and height were measured at the 6-week home visit and used to calculate maternal BMI (kg/m2). Although other demographic variables (including maternal age, maternal employment status, and maternal marital status) showed some correlations with primary study variables at individual time pints, none were significantly related to primary study variables in linear growth models. Additionally, primary results were not significantly changed by the inclusion of these additional covariates; therefore, in the interest of parsimony, these covariates were not included in final models.

Data Analysis

Primary analyses were conducted using Mplus v. 8. (Muthen & Muthen, 1998–2017) with maximum likelihood estimation to handle missing data. A linear growth model was estimated predicting the intercept and slope of child BMI from family economic hardship and neighborhood concentrated disadvantage, with duration of breastfeeding as a mediator. Maternal country of birth, maternal BMI, and child birth weight were included as covariates. Coefficients were scaled according to years from 12 months; the last coefficient was free to vary. Model fit was evaluated using comparative fit index (CFI) ≥ 0.95, root-mean-square residual (RMSEA) ≤ 0.08, and standardized root-mean-square residual (SRMR) < 0.08 (Hu & Bentler, 1999). Mediation was tested by examining the statistical significance of the indirect effect using 3000 bootstrap confidence intervals, which has been demonstrated to be a better statistical test of mediation than use of p values (MacKinnon, Lockwood, & Williams, 2004). If the 95% confidence interval does not include zero, the path is considered statistically significant (MacKinnon et al., 2004).

Results

Preliminary analyses.

Table 2 presents descriptive statistics and zero-order correlations among primary study variables. A base growth model evaluated an intercept and slope for children’s BMI, with no covariates or predictors. Both the intercept (Est = 1.980, SE Est = 0.312, p < .001) and slope (Est = 0.220, SE Est = 0.061, p < .001) had statistically significant variances, indicating sufficient variability among children to predict both initial 12-month BMI and BMI slope through 54 months. The mean of the intercept was significantly different from zero (Est = 15.157, SE Est = 1.393, p < .001). The mean of the slope was not significantly different from zero (Est = −0.068, SE Est = 0.814, p = .93). The intercept and slope were not significantly correlated (p = .14).

Table 2.

Descriptive Statistics and Correlations Among Primary Study Variables

Variable Mean (SD) Range 1 2 3 4 5 6 7 8 9 10
1. Economic hardship .01 (3.00) −6.47, 10.12 —
2. Neighborhood concentrated disadv. .00 (.73) −1.70, 2.61 −.009 —
3. Breastfeeding duration 1.70 (1.59) .00, 4.00 −.006 −.137a —
4. Child BMI (12 months) 18.12 (2.12) 12.94, 26.56 .041 .090 −.061 —
5. Child BMI (18 months) 17.01 (2.08) 12.37, 25.52 .051 .152 −.218 .456 —
6. Child BMI (24 months) 16.91 (1.86) 11.69, 23.35 .053 .247 −.218 .495 .580 —
7. Child BMI (36 months) 16.54 (1.86) 13.06, 24.07 .020 .188 −.163 .393 .495 .736 —
8. Child BMI (54 months) 16.50 (2.26) 12.69, 27.37 .042 .178 −.034 .375 .433 .579 .761 —
9. Country of birthb 1.86 (.34) 1, 2 .091 −.053 .290 .023 −.131 −.058 −.152 −.038 —
10. Child birth weight 3389 (464) 1190, 4935 .102 .057 .102 .210 .121 .135 .117 .233 .061 —
11. Maternal BMI 29.10 (5.84) 17.10, 50.00 .070 .026 .035 .211 .144 .159 .198 .297 −.060 .253
Open in a new tab
a

Bolded values indicate significant correlations at p < .05; “disadv.” = disadvantage.

b

Maternal country of birth, coded 1 = U.S., 2 = Mexico.

Potential outliers were identified by standardizing primary study variables; cases more than three standard deviations from the mean were evaluated further. One case was a potential outlier on family economic hardship, one case was a potential outlier on 12-month BMI, two cases were potential outliers on 18-month BMI, and one case a potential outlier on 24-month, 36-month, and 54-month BMI. The pattern of results did not change when these cases were excluded from analyses and thus all are included in subsequent analyses.

Prevalence of underweight, overweight, and obesity among children.

The prevalence of the current sample of children falling into underweight, overweight, and obesity categories at each time point was identified based on percentiles from gender and age norms (see Table 3; WHO, 2006). Across the five time points, the prevalence of overweight ranged from 13.5% to 18.1%, and the prevalence of obesity ranged from 18.8% to 34.2%.

Table 3.

Prevalence of Underweight, Overweight, and Obesity in the Sample of Children (Percentage)

Child age Nchildren Underweight Overweight Obese
12 months 187 2.1% 16.6% 34.2%
18 months 213 4.7% 17.4% 23.0%
24 months 216 3.2% 18.1% 23.1%
36 months 192 3.6% 13.5% 18.8%
54 months 203 2.0% 16.3% 19.7%
Open in a new tab

Note. Underweight is defined as a BMI that is below the fifth percentile, overweight as BMI in the 85th to 94th percentile, and obesity as BMI ≥ 95th percentile (in accordance with CDC standards; U.S. Preventive Services Task Force, 2010). Prevalence of underweight, overweight, and obese not presented according to IOTF definitions due to the age of the sample.

Prediction of children’s BMI.

A linear growth model was evaluated predicting child BMI intercept and slope from prenatal family economic hardship, prenatal neighborhood concentrated disadvantage, and breastfeeding duration (partial or exclusive vs. formula-feeding), adjusting for maternal country of birth, maternal BMI, and birth weight. The model demonstrated an adequate fit of the data: CFI = 0.912, RMSEA = 0.067 (90% CI: [0.047, 0.088]), SRMR = 0.104 (see Table 4 and Figure 1). Breastfeeding duration was significantly predicted by neighborhood concentrated disadvantage (Est = −0.275, SE Est = 0.109, p = .011) and maternal country of birth (Est = 1.314, SE Est = 0.179, p < .001), such that mothers living in more disadvantaged neighborhoods had a shorter duration of breastfeeding, but those born in Mexico breastfed for a longer duration than mothers born in the U.S. The intercept of child BMI was significantly predicted by neighborhood concentrated disadvantage (Est = 0.391, SE Est = 0.194, p = .044), breastfeeding duration (Est = −0.182, SE Est = 0.076, p = .017), and maternal BMI (Est = 0.050, SE Est = 0.024, p = .034). Living in a more disadvantaged prenatal neighborhood predicted higher child BMI, and a longer duration of breastfeeding predicted lower child BMI. However, none of the proposed variables were statistically significant predictors of the slope of child BMI.

Table 4.

Full Model Results

DV IV B SE B p R2 f2
Child BMI intercept 0.177* 0.22
Family economic hardship 0.018 0.041 0.657
Neighborhood concentrated disadvantage 0.391 0.194 0.044
Breastfeeding duration −0.182 0.076 0.017
Maternal country of birth −0.166 0.505 0.742
Child birth weight 0.050 0.026 0.056
Maternal BMI 0.050 0.024 0.017
Child BMI slope 0.020 0.02
Family economic hardship 0.003 0.025 0.894
Neighborhood concentrated disadvantage 0.009 0.112 0.935
Breastfeeding duration 0.055 0.046 0.233
Maternal country of birth −0.189 0.308 0.539
Child birth weight −0.011 0.017 0.530
Maternal BMI 0.004 0.014 0.774
Breastfeeding duration 0.109** 0.12
Family economic hardship −0.020 0.028 0.467
Neighborhood concentrated disadvantage −0.275 0.109 0.011
Maternal country of birth 1.314 0.179 0.000
Child birth weight 0.031 0.018 0.083
Open in a new tab
*

indicates statistically significant, p < .05.

**

indicates statistically significant, p < .001.

Figure 1.

Figure 1.

Open in a new tab

Growth model predicting children’s weight from family economic hardship, neighborhood concentrated disadvantage, breastfeeding, and covariates (maternal country of birth, maternal BMI, child birth weight). BMI = body mass index; Mat. = maternal; CD = concentrated disadvantage; Econ. = economic; BF = breastfeeding. Solid lines represent statistically significant pathways, and dashed lines are statistically nonsignificant. Error terms and covariances not shown.

Given the direct effects of neighborhood concentrated disadvantage on breastfeeding duration (a path), breastfeeding duration on the intercept of child BMI (b path), and neighborhood concentrated disadvantage on the intercept of child BMI (c′ path), the indirect effect of neighborhood concentrated disadvantage on child BMI via breastfeeding duration was evaluated. The specific indirect effect of neighborhood concentrated disadvantage on the intercept of child BMI through breastfeeding duration was statistically significant (Est = 0.050, SE = 0.029, 95% CI: [.002, .116]), suggesting breastfeeding duration partially mediates the effect of neighborhood concentrated disadvantage on the intercept of child BMI.

To assess the stability of the impact of prenatal neighborhood concentrated disadvantage on child BMI, the same linear growth model was reevaluated with coefficients scaled according to different intercepts (at 18, 24, 36, and 54 months). Prenatal neighborhood concentrated disadvantage was a significant predictor of child BMI intercept at 18 months (Est = 0.206, SE Est = 0.084, p = .015), 24 months (Est = 0.401, SE Est = 0.144, p = .006), 36 months (Est = 0.410, SE Est = 0.163, p = .012), and 54 months (Est = 0.479, SE Est = 0.222, p = .031).

The linear growth model was repeated with exclusive breastfeeding duration as a mediator in the prediction of child BMI intercept and slope from prenatal family economic hardship and prenatal neighborhood concentrated disadvantage, adjusting for maternal country of birth, maternal BMI, and birth weight. The pattern of results was comparable to the primary model, with an adequate fit of the data: CFI = 0.911, RMSEA = 0.066 (90% CI: [0.046, 0.086]), SRMR = 0.103. Neighborhood concentrated disadvantage (Est = 0.427, SE Est = 0.204, p = .036), exclusive breastfeeding duration (Est = −0.304, SE Est = 0.104, p = .003), maternal BMI (Est = 0.046, SE Est = 0.023, p = .047), child birth weight (Est = 0.051, SE Est = 0.025, p = .040), predicted child BMI intercept. However, the relation of neighborhood concentrated disadvantage to exclusive breastfeeding duration did not reach statistical significance (Est = −0.100, SE Est = 0.060, p = .10), precluding significant mediation. In line with prior analyses, none of the variables significantly predicted child BMI slope.

Adjustment for neighborhood clustering.

Despite 322 participants in the sample, there were only 152 unique neighborhoods (census tracts) at the prenatal timepoint. Therefore, the same linear growth model was evaluated adjusting for neighborhood clustering. Neighborhood concentrated disadvantage remained a significant predictor of breastfeeding duration (Est = −0.275, SE Est = .0.101, p = .006). Neighborhood concentrated disadvantage remained a significant predictor of the intercept of child BMI at 18 months (Est = 0.396, SE Est = .191, p = .038), 24 months (Est = 0.401, SE Est = 0.181, p = .027), 36 months (Est = 0.439, SE Est = 0.178, p = .013), and 54 months (Est = 0.479, SE Est = 0.234, p = .041). Neighborhood concentrated disadvantage was a marginally significant predictor of the intercept of child BMI at 12 months (Est = 0.391, SE Est = 0.214, p = .068).

Neighborhood concentrated disadvantage and BMI classification.

Table 5 displays estimated child BMI at one standard deviation (SD) above and below the mean on neighborhood concentrated disadvantage at each time point. These BMI values were compared to WHO Child Growth Standards (WHO, 2006) using the WHO Anthropometric Calculator to determine whether they fell into underweight, healthy, overweight, or obese categorization. At one SD above the mean, indicating higher disadvantage, the estimated BMI fell into the overweight category at all time points.

Table 5.

BMI and CD

Neighborhood concentrated disadvantage
BMI −1 SD +1 SD
12 months 17.95* 18.31**
18 months 16.69 17.31**
24 months 16.43 17.38*
36 months 16.20 16.88*
54 months 16.12 16.92**
Open in a new tab
*

indicates BMI falls in overweight classification (85th to 94th percentile) or borderline overweight classification for girls.

**

indicates BMI falls in overweight classification (85th to 94th percentile) or borderline overweight classification for girls and boys.

Discussion

Longitudinal studies have the potential to identify prenatal and early life factors that contribute to the development of childhood obesity and contribute to ethnic minority health disparities. The current study evaluated the direct and indirect (via breastfeeding duration) influences of prenatal family- and neighborhood-level disadvantage on children’s weight gain from infancy through age 4.5 years among a sample of low-income, Mexican American families. Prenatal neighborhood disadvantage emerged as a statistically significant predictor of child BMI in our model, above and beyond family socioeconomic hardship and several well-documented predictors of children’s weight (maternal country of birth, birth weight, maternal BMI). The relation of neighborhood disadvantage to child BMI was partially mediated by a shorter duration of partial or exclusive breastfeeding among women living in more disadvantaged neighborhoods. Breastfeeding duration may be a modifiable pathway through which neighborhood disadvantage impacts children’s risk of unhealthy weight status.

Contrary to hypotheses, prenatal family-level economic disadvantage was not associated with breastfeeding duration or children’s BMI. Exposure to family-level poverty throughout development is thought to be influential to children’s weight and risk of obesity for various reasons. A lack of resources most directly affects the household obesogenic environment and children’s weight trajectories through food insecurity, or a family’s inability to attain nutritionally appropriate foods due to financial, social, or physical barriers (Anderson, 1990). Limited work has focused specifically on exposure to family economic conditions during the prenatal period in studies of children’s weight status. However, existing work suggests earlier exposure to family economic disadvantage to be most influential. Ziol-Guest and colleagues (2009) found family income during the earliest period of life (prenatal and birth years) to be the strongest predictor of increased adult BMI; family economic conditions during one and five years of age and between six and 15 years of age were not associated with later BMI. The impact of family-level economic disadvantage on children’s weight may be more evident in later childhood, adolescence, or adulthood. However, many studies examining the impact of family economic conditions on children’s weight status fail to simultaneously account for broader contexts, like neighborhoods, that may overshadow the effects of family economic hardship.

As hypothesized, maternal residence in a neighborhood with higher concentrated disadvantage during the prenatal period predicted higher infant BMI at 12 months. This effect persisted through child age 4.5 years (at 18, 24, 36, and 54 months). More specifically, maternal residence in a disadvantaged neighborhood one or more SD above the sample mean was associated with child BMI in the overweight classification through 4.5 years, whereas residence in less disadvantaged neighborhoods (one or more SD below the mean) was associated with BMI in the normal range from ages 1–4.5. However, at 12 months, living in a less disadvantaged neighborhood (one SD below the mean) is associated with borderline overweight status in children, demonstrating that this sample of children is particularly at risk for unhealthy weight. Although the magnitude of difference in the estimated BMI at one SD above and below the mean on neighborhood concentrated disadvantage appeared relatively small (less than 1 kg/m2 at all time points), it is important to consider the early age at which it is apparent. Additionally, the sample as a whole experienced significant economic hardship and resided in neighborhoods that were disadvantaged relative to the broader metropolitan area. The current results underscore how the degree of disadvantage has a significant effect on obesity risk evident early in life, even among a more broadly disadvantaged population.

Although this study did not examine specific aspects of the prenatal neighborhood, previous research suggests potential factors that may operate during the prenatal period to negatively impact child obesity risk. For example, maternal stress and limited availability of nutritious foods may link prenatal neighborhood disadvantage with childhood obesity risk (Anderson, 1990; Cook & Frank, 2008; Walker, Keane, & Burke, 2010). Women may become more aware of their environmental context during the perinatal period (Hahn-Holbrook, Holbrook, & Haselton, 2011), and perhaps more specifically experience a heightened awareness of the negative aspects of the neighborhood, thus increasing maternal stress. Pregnancy may be an ecological transition point in women’s lives where chronic stressors such as neighborhood disadvantage become more salient and influential. Future work may aim to uncover the specific mechanisms underlying the relation between prenatal neighborhood disadvantage and childhood obesity risk.

The current study evaluated the duration of partial or exclusive breastfeeding as a potential early life mechanism explaining the association between family- and neighborhood-level disadvantage and children’s obesity risk. Although prenatal family economic disadvantage did not predict the duration of partial or exclusive breastfeeding, greater neighborhood disadvantage was associated with a shorter duration of partial or exclusive breastfeeding. In addition, breastfeeding duration was significantly associated with children’s BMI at all time points. The duration of partial or exclusive breastfeeding was a partial mediator between neighborhood disadvantage and children’s BMI. The duration of exclusive breastfeeding was not a statistically significant mediator, suggesting that the small percentage of women who exclusively breastfeed may be less impacted by characteristics of the neighborhood. Importantly, breastfeeding is a modifiable risk factor that could potentially mitigate the negative impact of prenatal neighborhood disadvantage on early life obesity risk. Contemporary theory suggests that neighborhood-level social support and educational resources available for breastfeeding within a community can contribute to the likelihood and duration of breastfeeding (Burdette, 2013). Breastfeeding is particularly important among families experiencing socioeconomic hardship. Underserved mothers are at risk for numerous postpartum difficulties such as obesity (Flegal, Carroll, Ogden, & Curtin, 2010), type-2 diabetes (CDC, 2011), and unintended pregnancies (Department of Health and Human Services), but if breastfeeding, experience healthier postpartum weight loss (Baker et al., 2008), less type-2 diabetes (Stuebe et al., 2011), and longer intervals between offspring (Kennedy & Visness et al., 1992). Public health media campaigns and community-level resources dedicated to increasing the rates and duration of breastfeeding among families living in disadvantaged environments may reduce the long-term societal costs of childhood obesity.

Maternal country of birth emerged as a significant predictor of breastfeeding duration, such that women born in the United States reported shorter duration of partial or exclusive breastfeeding. Cultural considerations may be especially relevant to public health programs targeting Mexican-origin women in the United States. In the current study, maternal country of birth may serve as a proxy measure of maternal acculturation, and a large body of literature suggests that greater acculturation to mainstream U.S. culture is associated with decreased likelihood of initiating and sustaining breastfeeding among Hispanic women (Bigman, Wilkinson, Pérez, & Homedes, 2018). Future work may aim to understand the role of the neighborhood cultural context as it relates to breastfeeding behaviors.

Although higher neighborhood concentrated disadvantage, longer breastfeeding duration, and higher maternal postpartum BMI were predictive of higher child BMI at 12 months, none of the variables in our model were associated with the slope of children’s BMI through 4.5 years. In addition to the first year of life, alternative critical periods for the development of obesity during childhood include the beginning of school and adiposity rebound (Adair, 2008). In Hispanic children, adiposity rebound typically occurs around age 5 to 5.5 (Boonpleng, Park, & Gallo, 2012). Perhaps, the current study’s evaluation of BMI change through age 4.5 was too early to capture whether prenatal disadvantage and breastfeeding duration impact the slope of BMI during this second influential period. Predictors of early life weight may be different from those that predict subsequent weight gain.

The current findings should be interpreted considering several limitations. First, the sample consisted of low-income, Mexican American families and findings may not generalize to other groups. Second, the dependent variable in this study is BMI, a proxy measure for obesity. Research has suggested that BMI is a useful screener for obesity but can yield varied measurement error (Rothman, 2008). Mothers were interviewed to assess breastfeeding information and family economic hardship, both of which can be susceptible to social desirability. Additionally, our data collection distinguished breastfeeding from bottle-feeding, but did not address the distinction between formula and bottled breast milk. Our measure also precluded the ability to differentiate between levels of partial breastfeeding (e.g., high vs. medium vs. low; Labbok & Krasovec, 1990), which may have important implications for the development of prevention and intervention programs with this vulnerable population. Relatedly, results from additional analyses of exclusive breastfeeding should be interpreted with caution given the small number of women who exclusively breastfed across six months. This lack of power may explain why evidence of mediation was found using partial or exclusive breastfeeding duration, but not solely exclusive breastfeeding duration. Still, the estimates of both mediated paths are in the same hypothesized direction. Finally, the current study cannot account for the effects of moving to higher or lower advantaged neighborhoods after the prenatal visit. Neighborhoods are thought to continuously and concurrently influence pathways of health, however the current study assumes the prenatal neighborhood to be independently influential to subsequent health outcomes.

The current study’s unique, longitudinal perspective beginning prenatally allows for an understanding of children’s weight status during early, critical years and contributes to a growing body of literature implicating the prenatal period as particularly influential. Although low-income, Mexican American children are at an elevated risk for overweight and obesity, less is known about the familial and environmental factors that drive obesity risk among this population. In this sample of low-income, Mexican American children, a high prevalence of overweight and obesity was found from 12 months to 4.5 years, indicative of an elevated risk for obesity throughout development. By evaluating the effects of economic disadvantage across two systemic levels, families and neighborhoods, the study revealed that the prenatal neighborhood context may be especially important for children’s weight status during the early years, over and above the family economic context. Furthermore, breastfeeding duration, a modifiable health behavior, emerged as one pathway through which the prenatal neighborhood impacts children’s BMI. Among this high-risk population, community-level preventive efforts targeting breastfeeding may be beneficial for combating unhealthy weight status during early childhood.

Acknowledgments

The study was funded by the National Institute of Mental Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Mental Health. We thank the mothers and infants for their participation; Kirsten Letham, Monica Gutierrez, Elizabeth Nelson, and Jody Southworth-Brown for their assistance with data collection and management; Dr. Dean Coonrod and the Maricopa Integrated Health System for their assistance with recruitment; and the interviewers for their commitment and dedication to this project.

References

  1. Adair LS (2008). Child and adolescent obesity: Epidemiology and developmental perspectives. Physiology & Behavior, 94, 8–16. 10.1016/j.physbeh.2007.11.016 [DOI] [PubMed] [Google Scholar]
  2. Alvarado R, Zepeda A, Rivero S, Rico N, López S, & Díaz S (1999). Integrated maternal and infant health care in the postpartum period in a poor neighborhood in Santiago, Chile. Studies in Family Planning, 30, 133–141. 10.1111/j.1728-4465.1999.00133.x [DOI] [PubMed] [Google Scholar]
  3. Anderson PM, & Butcher KE (2006). Childhood obesity: Trends and potential causes. The Future of Children, 16, 19–45. 10.1353/foc.2006.0001 [DOI] [PubMed] [Google Scholar]
  4. Anderson SA (1990). Core indicators of nutritional state for difficult-to-sample populations. The Journal of Nutrition, 120, 1555–1600. 10.1093/jn/120.suppl_11.1555 [DOI] [PubMed] [Google Scholar]
  5. Anderson SE, & Whitaker RC (2009). Prevalence of obesity among U.S. preschool children in different racial and ethnic groups. Archives of Pediatrics & Adolescent Medicine, 163, 344–348. 10.1001/archpediatrics.2009.18 [DOI] [PubMed] [Google Scholar]
  6. Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, & Law C (2005). Being big or growing fast: Systematic review of size and growth in infancy and later obesity. British Medical Journal (Clinical Research Ed.), 331, 929. 10.1136/bmj.38586.411273.E0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Baker JL, Gamborg M, Heitmann BL, Lissner L, Sørensen TI, & Rasmussen KM (2008). Breastfeeding reduces postpartum weight retention. The American Journal of Clinical Nutrition, 88, 1543–1551. 10.3945/ajcn.2008.26379 [DOI] [PubMed] [Google Scholar]
  8. Barrera M Jr., Caples H, & Tein JY (2001). The psychological sense of economic hardship: Measurement models, validity, and cross-ethnic equivalence for urban families. American Journal of Community Psychology, 29, 493–517. 10.1023/A:1010328115110 [DOI] [PubMed] [Google Scholar]
  9. Bickel G, Andrews M, & Klein B (1996). Measuring food security in the United States: a supplement to the CPS. Nutrition and Food Security in the Food Stamp Program, 91–111. [Google Scholar]
  10. Bigman G, Wilkinson AV, Pérez A, & Homedes N (2018). Acculturation and breastfeeding among Hispanic American women: A systematic review. Maternal and Child Health Journal, 22, 1260–1277. 10.1007/s10995-018-2584-0 [DOI] [PubMed] [Google Scholar]
  11. Biro FM, & Wien M (2010). Childhood obesity and adult morbidities. The American Journal of Clinical Nutrition, 91, 1499S–1505S. 10.3945/ajcn.2010.28701B [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Boonpleng W, Park CG, & Gallo AM (2012). Timing of adiposity rebound: A step toward preventing obesity. Pediatric Nursing, 38, 37–42. Retrieved from http://www.pediatricnursing.net/index.html [PubMed] [Google Scholar]
  13. Bunik M, Clark L, Zimmer LM, Jimenez LM, O’Connor ME, Crane LA, & Kempe A (2006). Early infant feeding decisions in low-income Latinas. Breastfeeding Medicine, 1, 225–235. 10.1089/bfm.2006.1.225 [DOI] [PubMed] [Google Scholar]
  14. Burdette AM (2013). Neighborhood context and breastfeeding behaviors among urban mothers. Journal of Human Lactation, 29, 597–604. 10.1177/0890334413495110 [DOI] [PubMed] [Google Scholar]
  15. Cao-Lei L, Laplante DP, & King S (2016). Prenatal maternal stress and epigenetics: Review of the human research. Current Molecular Biology Reports, 2, 16–25. 10.1007/s40610-016-0030-x [DOI] [Google Scholar]
  16. Carroll-Scott A, Gilstad-Hayden K, Rosenthal L, Peters SM, McCaslin C, Joyce R, & Ickovics JR (2013). Disentangling neighborhood contextual associations with child body mass index, diet, and physical activity: The role of built, socioeconomic, and social environments. Social Science & Medicine, 95, 106–114. 10.1016/j.socscimed.2013.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Centers for Disease Control and Prevention. (2011). National diabetes fact sheet: National estimates and general information on diabetes and prediabetes in the United States, 2011. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Retrieved from https://www.cdc.gov/chronicdisease/center/news-media/archives/publications.htm [Google Scholar]
  18. Centers for Disease Control and Prevention (CDC). (2010). Racial and ethnic differences in breastfeeding initiation and duration, by state—National Immunization Survey, United States, 2004–2008. Morbidity and Mortality Weekly Report, 59, 327–334. Retrieved from https://www.cdc.gov/chronicdisease/center/news-media/archives/publications.htm [PubMed] [Google Scholar]
  19. Chen E, & Miller GE (2013). Socioeconomic status and health: Mediating and moderating factors. Annual Review of Clinical Psychology, 9, 723–749. 10.1146/annurev-clinpsy-050212-185634 [DOI] [PubMed] [Google Scholar]
  20. Chung A, Backholer K, Wong E, Palermo C, Keating C, & Peeters A (2016). Trends in child and adolescent obesity prevalence in economically advanced countries according to socioeconomic position: A systematic review. Obesity Reviews, 17, 276–295. 10.1111/obr.12360 [DOI] [PubMed] [Google Scholar]
  21. Conroy K, Sandel M, & Zuckerman B (2010). Poverty grown up: How childhood socioeconomic status impacts adult health. Journal of Developmental and Behavioral Pediatrics, 31, 154–160. 10.1097/DBP.0b013e3181c21a1b [DOI] [PubMed] [Google Scholar]
  22. Cook JT, & Frank DA (2008). Food security, poverty, and human development in the United States. Annals of the New York Academy of Sciences, 1136, 193–209. 10.1196/annals.1425.001 [DOI] [PubMed] [Google Scholar]
  23. Cunradi CB, Caetano R, Clark C, & Schafer J (2000). Neighborhood poverty as a predictor of intimate partner violence among White, Black, and Hispanic couples in the United States: A multilevel analysis. Annals of Epidemiology, 10, 297–308. 10.1016/S1047-2797(00)00052-1 [DOI] [PubMed] [Google Scholar]
  24. Daoud N, O’Campo P, Minh A, Urquia ML, Dzakpasu S, Heaman M, … Chalmers B (2014). Patterns of social inequalities across pregnancy and birth outcomes: A comparison of individual and neighborhood socioeconomic measures. BMC Pregnancy and Childbirth, 14, 393. 10.1186/s12884-014-0393-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Dubowitz T, Ghosh-Dastidar M, Eibner C, Slaughter ME, Fernandes M, Whitsel EA, … Escarce JJ (2012). The Women’s Health Initiative: The food environment, neighborhood socioeconomic status, BMI, and blood pressure. [Silver Spring]. Obesity, 20, 862–871. 10.1038/oby.2011.141 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Fang D, Thomsen MR, Nayga RM Jr., & Goudie A (2018). Association of neighborhood geographic spatial factors with rates of childhood obesity. Journal of the American Medical Association Network Open, 1(4), e180954–e180954. 10.1001/jamanetworkopen.2018.0954 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Flegal KM, Carroll MD, Ogden CL, & Curtin LR (2010). Prevalence and trends in obesity among U.S. adults, 1999–2008. Journal of the American Medical Association, 303, 235–241. 10.1001/jama.2009.2014 [DOI] [PubMed] [Google Scholar]
  28. Glass TA, & Bilal U (2016). Are neighborhoods causal? Complications arising from the “stickiness” of ZNA. Social Science & Medicine, 166, 244–253. 10.1016/j.socscimed.2016.01.001 [DOI] [PubMed] [Google Scholar]
  29. Guerrero AD, Mao C, Fuller B, Bridges M, Franke T, & Kuo AA (2016). Racial and ethnic disparities in early childhood obesity: Growth trajectories in body mass index. Journal of Racial and Ethnic Health Disparities, 3, 129–137. 10.1007/s40615-015-0122-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hahn-Holbrook J, Holbrook C, & Haselton MG (2011). Parental precaution: Neurobiological means and adaptive ends. Neuroscience and Biobehavioral Reviews, 35, 1052–1066. 10.1016/j.neubiorev.2010.09.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hales CM, Carroll MD, Fryar CD, & Ogden CL (2017). Prevalence of obesity among adults and youth: United States, 2015–2016. Retrieved from https://www.cdc.gov/chronicdisease/center/news-media/archives/publications.htm [PubMed]
  32. Hamilton WL, & Cook JT (1997). Household food security in the United States in 1995: technical report of the food security measurement project.
  33. Hediger ML, Overpeck MD, Ruan WJ, & Troendle JF (2000). Early infant feeding and growth status of US-born infants and children aged 4–71 mo: analyses from the third National Health and Nutrition Examination Survey, 1988–1994. The American journal of clinical nutrition, 72, 159–167. 10.1093/ajcn/72.1.159 [DOI] [PubMed] [Google Scholar]
  34. Hernández-Valero MA, Wilkinson AV, Forman MR, Etzel CJ, Cao Y, Bárcenas CH, … Bondy ML (2007). Maternal BMI and country of birth as indicators of childhood obesity in children of Mexican origin. Obesity, 15, 2512–2519. 10.1038/oby.2007.298 [DOI] [PubMed] [Google Scholar]
  35. Horta BL, & Victora CG (2013). Long-term effects of breastfeeding. Geneva, Switzerland: World Health Organization. Retrieved from https://www.who.int/childgrowth/publications/en/ [Google Scholar]
  36. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  37. Inagami S, Cohen DA, Brown AF, & Asch SM (2009). Body mass index, neighborhood fast food and restaurant concentration, and car ownership. Journal of Urban Health, 86, 683–695. 10.1007/s11524-009-9379-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Kaufman L, Deenadayalan S, & Karpati A (2010). Breastfeeding ambivalence among low-income African American and Puerto Rican women in north and central Brooklyn. Maternal and Child Health Journal, 14, 696–704. 10.1007/s10995-009-0499-5 [DOI] [PubMed] [Google Scholar]
  39. Kennedy KI, & Visness CM (1992). Contraceptive efficacy of lactational amenorrhoea. The Lancet, 339, 227–230. 10.1016/0140-6736(92)90018-X [DOI] [PubMed] [Google Scholar]
  40. Kids Count Data Center. (2017). Children in poverty by race and ethnicity. Retrieved from https://datacenter.kidscount.org/data/tables/44-children-in-poverty-by-race-and-ethnicity
  41. Kids Count Data Center. (2019). Children living in high poverty areas by race and ethnicity in the United States. Retrieved from https://datacenter.kidscount.org/data/tables/7753-children-living-in-areas-of-concentrated-poverty-by-race-and-ethnicity#detailed/1/any/false/1691,1607,1572,1485,1376,1201,1074,880/10,11,9,12,1,185,13/,14943,14942
  42. King KE, Kane JB, Scarbrough P, Hoyo C, & Murphy SK (2016). Neighborhood and family environment of expectant mothers may influence prenatal programming of adult cancer risk: Discussion and an illustrative DNA methylation example. Biodemography and Social Biology, 62, 87–104. 10.1080/19485565.2015.1126501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Labbok M, & Krasovec K (1990). Toward consistency in breastfeeding definitions. Studies in Family Planning, 21, 226–230. 10.2307/1966617 [DOI] [PubMed] [Google Scholar]
  44. Leclair E, Robert N, Sprague AE, & Fleming N (2015). Factors associated with breastfeeding initiation in adolescent pregnancies: A cohort study. Journal of Pediatric and Adolescent Gynecology, 28, 516–521. 10.1016/j.jpag.2015.03.007 [DOI] [PubMed] [Google Scholar]
  45. MacKinnon DP, Lockwood CM, & Williams J (2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39, 99–128. 10.1207/s15327906mbr3901_4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Muthén L, & Muthén B (2017). Mplus (Version 8) [Computer software]. Los Angeles, CA: Author. Retrieved from https://www.statmodel.com/ [Google Scholar]
  47. Nicholas LM, Morrison JL, Rattanatray L, Zhang S, Ozanne SE, & McMillen IC (2016). The early origins of obesity and insulin resistance: Timing, programming and mechanisms. International Journal of Obesity, 40, 229–238. 10.1038/ijo.2015.178 [DOI] [PubMed] [Google Scholar]
  48. Ogden CL, Carroll MD, Kit BK, & Flegal KM (2012). Prevalence of obesity and trends in body mass index among U.S. children and adolescents, 1999–2010. Journal of the American Medical Association, 307, 483–490. 10.1001/jama.2012.40 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Owen CG, Martin RM, Whincup PH, Smith GD, & Cook DG (2005). Effect of infant feeding on the risk of obesity across the life course: A quantitative review of published evidence. Pediatrics, 115, 1367–1377. 10.1542/peds.2004-1176 [DOI] [PubMed] [Google Scholar]
  50. Rothman KJ (2008). BMI-related errors in the measurement of obesity. International Journal of Obesity, 32(Suppl. 3), S56–S59. 10.1038/ijo.2008.87 [DOI] [PubMed] [Google Scholar]
  51. Saelens BE, Glanz K, Frank LD, Couch SC, Zhou C, Colburn T, & Sallis JF (2018). Two-year changes in child weight status, diet, and activity by neighborhood nutrition and physical activity environment. Obesity, 26, 1338–1346. 10.1002/oby.22247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sampson RJ, Raudenbush SW, & Earls F (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918–924. 10.1126/science.277.5328.918 [DOI] [PubMed] [Google Scholar]
  53. Schulz A, & Northridge ME (2004). Social determinants of health: Implications for environmental health promotion. Health Education & Behavior, 31, 455–471. 10.1177/1090198104265598 [DOI] [PubMed] [Google Scholar]
  54. Shealy KR, Li R, Benton-Davis S, & Grummer-Strawn LM (2005). The CDC guide to breastfeeding interventions. Atlanta, Georgia: US Department of Health and Human Services, Centers for Disease Control and Prevention. [Google Scholar]
  55. Shonkoff JP, & Phillips DA (2000). From neurons to neighborhoods: The science of early childhood development. Washington, DC: National Academies Press. 10.17226/9824 [DOI] [PubMed] [Google Scholar]
  56. Silverman JG, Decker MR, Reed E, & Raj A (2006). Intimate partner violence around the time of pregnancy: Association with breastfeeding behavior. Journal of Women’s Health, 15, 934–940. 10.1089/jwh.2006.15.934 [DOI] [PubMed] [Google Scholar]
  57. Singh GK, Siahpush M, & Kogan MD (2010). Neighborhood socioeconomic conditions, built environments, and childhood obesity. Health Affairs, 29, 503–512. 10.1377/hlthaff.2009.0730 [DOI] [PubMed] [Google Scholar]
  58. Singhal A, & Lanigan J (2007). Breastfeeding, early growth and later obesity. Obesity Reviews, 8(Suppl. 1), 51–54. 10.1111/j.1467-789X.2007.00318.x [DOI] [PubMed] [Google Scholar]
  59. Stuebe AM, Schwarz EB, Grewen K, Rich-Edwards JW, Michels KB, Foster EM, … Forman J (2011). Duration of lactation and incidence of maternal hypertension: A longitudinal cohort study. American Journal of Epidemiology, 174, 1147–1158. 10.1093/aje/kwr227 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. U.S. Census Bureau. (2011). 2010 American Community Survey 1-year estimates. Retrieved from http://factfinder2.census.gov
  61. U.S. Preventive Services Task Force. (2010). Screening for obesity in children and adolescents: US Preventive Services Task Force recommendation statement. Pediatrics, 125, 361–367. [DOI] [PubMed] [Google Scholar]
  62. Walker RE, Keane CR, & Burke JG (2010). Disparities and access to healthy food in the United States: A review of food deserts literature. Health & Place, 16, 876–884. 10.1016/j.healthplace.2010.04.013 [DOI] [PubMed] [Google Scholar]
  63. WHO Multicentre Growth Reference Study Group. (2006). WHO Child Growth Standards based on length/height, weight and age. Acta Paediatrica, 450, 76. Retrieved from https://www.who.int/childgrowth/publications/en/16817681 [Google Scholar]
  64. World Health Organization (WHO). (2006). WHO Child Growth Standards: Length/height-for age, weight-for-age, weight for length, weight for height: Methods and development. Geneva, Switzerland: World Health Organization. Retrieved from https://www.who.int/childgrowth/publications/en/ [Google Scholar]
  65. Yourkavitch J, Kane JB, & Miles G (2018). Neighborhood disadvantage and neighborhood affluence: Associations with breastfeeding practices in urban areas. Maternal and Child Health Journal, 22, 546–555. 10.1007/s10995-017-2423-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Ziol-Guest KM, Duncan GJ, & Kalil A (2009). Early childhood poverty and adult body mass index. American Journal of Public Health, 99, 527–532. 10.2105/AJPH.2007.130575 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES