Comparison of Early Life Obesity-Related Risk and Protective Factors in Non-Hispanic Black Subgroups

Abstract

Objectives.

Previous obesity prevention studies in preschool-age children have included non-Hispanic Black (NHB) children, but few have investigated between-subgroup differences even though there may be cultural risk and protective practice differences, challenging the generalizability of findings. The purpose of this study was to examine differences in early childhood obesity-related factors in NHB subgroups (Haitian, other Caribbean Islander and African-American [AA]) children.

Methods.

Baseline data from two randomized controlled trials in 52 childcare centers of which 35 had data to test a preschool-based obesity prevention intervention was analyzed. The sub-sample included 370 caregiver-child dyads; 209 self-identified as AA, 120 as Haitian and 41 as Caribbean Islander/West Indian or mixed race. Multilevel regression models generated outcome estimates for group differences in body mass index (BMI) percentile, birthweight, breastfeeding initiation and duration, bottle feeding duration and age when solid foods were introduced.

Results.

Mean BMI percentile was similar for AA, Haitian and Caribbean Islander/West Indian/Multiracial (60.1^th percentile, 60.8^th percentile, 62.8^th percentile, respectively) as was birthweight (6.3, 6.8, and 6.6 pounds, respectively). Children of US-born caregivers had significantly lower BMI percentiles (9.13 percentile points) versus foreign-born caregivers. Haitian women were significantly more likely to initiate breastfeeding (64.9%) versus AA (47.6%) and Caribbean Islander/West Indian/Multiracial (62.2%)(p<.01). No significant group differences were found in breastfeeding or bottle feeding duration or age solid foods were introduced.

Conclusions.

Findings here suggest that NHB race classification can identify important subgroup behavioral similarities which in turn may inform culturally sensitive strategies to promote early childhood healthy weight. Foreign-born caregivers may benefit from healthy weight promotion information, and as early as possible in their child’s development.

Introduction

United States (US) population-level data shows that children from ethnic minority backgrounds are disproportionately affected by early-onset obesity compared to their non-ethnic group counterparts (Ogden et al. 2018). Specifically, non-Hispanic blacks (NHB) have higher rates of obesity (body mass index [BMI] ≥ 95^th percentile for age and sex) by age 5 compared to non-Hispanic whites (NHW) and Asians (11.6%, 9.9%, and 7.0%, respectively) (Ogden et al., 2018). Using National Health and Nutrition Examination Survey (NHANES) population-level data, analyses showed that among children as young as preschool-age, greater BMI and waist circumference (WC) are associated with cardiovascular disease risk biomarkers, and these associations vary by ethnicity (Messiah et al., 2012; Messiah et al., 2008).

Although racial/ethnic disparities in childhood obesity prevalence estimates and risk behaviors have been identified (Barnes et al, 2011; Perrin et al., 2014; Kimbro et al, 2007; Taveras et al., 2010; Lawrence et al. 2008; Dennison et al., 2002), the majority of studies included non-Hispanic white, high-income study samples, which decrease generalizability of conclusions. Furthermore, studies that do include racial/ethnic minorities are limited by panethnic umbrella terms like non-Hispanic black or African American (Nuru et al., 2018). Terms such as these do not acknowledge subgroup differences (e.g., among African American, Haitian, other Caribbean/West Indian immigrants, those who identify with more than one NHB subgroup or are mixed/multirace- ethnicity). One study examined obesity in US-born versus foreign-born NHBs and found that those who were foreign-born generally had lower obesity levels (Mehta et al., 2015). It is impossible to disentangle subgroup differences in these conclusions, however. Indeed, utilizing the generic ‘Black’ or ‘NHB’ racial classification “overlooks distinct within-group cultural norms and health behaviors that may influence risks for unhealthy weight” (Arthur & Katkin, 2006).

Florida, and South Florida in particular, is home to the majority (20%, equal to 1.1 million) of all US immigrants from the Caribbean (Kent, 2007). The Caribbean is in an epidemiological transition whereby non-communicable nutrition-related deaths are replacing deaths by communicable and infectious diseases (Henry, 2016). One study found that the gradient in per capita gross national product in three Caribbean island countries parallels the gradient in the proportions of population in those countries who are obese (Forrester et al., 2007), suggesting that Caribbean immigrant children from different countries of origin are exposed to different risks for disease. However, given the scarcity of research disaggregating NHB subgroups, and the lack of research examining and/or comparing NHB subgroups, these children may not receive appropriate, culturally specific or targeted health promotion and prevention efforts (Dietz, 2015; Brown, 2015; Seo & Jaesin, 2010).

Haitians are now the nation’s second largest Caribbean immigrant group in the US and have grown fourfold over the past 4 decades with the most recent statistics estimating there are approximately 600,000 Haitians living in the United States (Kent, 2007). For example, one study reported that youth ages 2–18 who were born in Haiti increased their body mass index (BMI) percentile by 3.7% for each year of US residency (Strickman-Stein et al., 2010). Although Haitian-born children were less likely to be an unhealthy weight compared to U.S.-born Haitian children, they were just as likely to be obese as their NHB and NHW peers suggesting a rapid assimilation to US culture.

This study explores the relations among early childhood obesity-related factors and BMI by caregiver-identified NHB subgroup caregiver-child dyads. We sought to explore the relationships among caregiver characteristics (education level, relationship to child, language spoken at home, and birthplace), early childhood behaviors (breastfeeding initiation and duration in months, bottle feeding use in months, age at early introduction of solid food) and BMI percentile among preschool-age children.

Methods

Main Protocol: Study Design

Data from this cross-sectional study were extracted from two large randomized controlled trials of ‘Healthy Caregivers, Healthy Children (HC2, Phase 1 and Phase 2, respectively), a randomized controlled intervention trial (Clinical Trial # NCT01722032 and # NCT02697565, respectively) for healthy weight maintenance among children 2-to-5 years old. Phase 1 was conducted in 28 subsidized childcare centers in Miami-Dade County, Florida between 2010 and 2013 and Phase 2 was conducted in 24 subsidized childcare centers in 2015–2018. The data for the analysis presented here combined measures from the baseline timepoint in both trials. This study protocol was reviewed and approved by the University of Miami Institutional Review Board. Each child’s parent or legal guardian provided informed consent.

Participants

Childcare Centers.

The following selection criteria were used to determine childcare center eligibility for both HC2 trials: (1) more than 30 2-to-5 year olds attending the center; (2) ethnicity reflective of Miami-Dade County’s population; (3) serve low-income families; and (4) center director agreement to participate. Centers were excluded if they did not meet these criteria. Centers with a high prevalence of special needs children were excluded. Special needs did not include children with food allergies and sensitivities. These children were included if their parent/caregiver consented to study participation. If children brought their own meals due to diet restrictions, they were excluded. Children with extreme allergies were also a part of the exclusion criteria but both trials found that these children did not participate in the childcare settings included in both trials. Children identified by parents on the demographic form as failure to thrive (b5th BMI %ile) were excluded (Messiah et al., 2017).

In HC2 Phase 1, 28 centers and in Phase 2 30 centers agreed to participate. These center directors and teachers were invited to attend a project orientation describing the HC2 activities at their centers. At that time all centers reinstated commitment and signed a consent form to participate. Staff met with the center directors to establish an individual plan that details the way the center would be involved in this HC2 project.

Parent/Caregiver Participants.

Parents/guardians were recruited for participation in both trials via verbal announcements, flyers, and letters sent home. Family members who returned a signed interest form were then contacted by telephone by a study staff member. After verbal consent was obtained, an appointment was made for the child and parent to be seen at the childcare center, at which time the written informed consent from the parents was obtained and initial baseline data are collected.

Secondary Data Analysis Participants.

The present analysis included only baseline data from children and their caregivers who identified as either African American, Haitian or “other category” that included non-Hispanic Caribbean Islander/West Indian, African, and/or multiracial. Across centers, an average of 70% of eligible families with children 2-to-5 years old participated in both trials. For this analysis if there were siblings, only the oldest child was included.

Procedures

Participants were recruited from the childcare center that they attended. Parents were approached during drop off or pick up times. Consent forms were attached to the interview packets, and parent data were collected during the initial visit. If parents were not available for a face-to-face interview, staff members scheduled a time to conduct the interview over the telephone. Interviewers were either African American or from a variety of countries (Cuba, Dominican Republic, Venezuela, Nicaragua, Puerto Rico, Haiti, Jamaica) to be representative of the Miami Dade County population. Interviewers were available who spoke English, Spanish and Creole. Once consent forms were collected, research assistants returned to the center within two weeks to collect each child’s height and weight which was then converted to an age-and sex-adjusted BMI percentile (Kuczmarski et al., 2000). All research assistants were trained in interview administration and anthropometric data collection. This training consisted of 2-hour seminars across multiple days involving lectures, demonstrations, and practice, and in vivo observation of interview techniques and ongoing supervision.

All variables included in this post hoc analysis were included in the “HC2 Parent Intake Form” which was the first set of data collection forms in the survey battery in both trials. This form included all key demographic variables (age, sex, race/ethnicity) as well as other parent characteristics (education level, language spoken at home, born in/outside of the US). It also included all early childhood feeding behaviors used in this analysis (see below for details).

Measures

Demographic Measures.

Baseline assessment of child age, sex, and ethnicity were collected at the first interview or survey collection timepoint. Research assistants asked caregivers “What sex is your child?”, “What age is your child?” and “What ethnic group do you consider this child to belong to?” Caregivers had the choice of one of the following racial/ethnic groups: Hispanic/Cuban, Hispanic/Mexican, Hispanic/Puerto Rican, Hispanic/Other, Haitian, Non-Hispanic Black, Non-Hispanic White, Other Caribbean Islander, Other.” Baseline measures of the following caregiver variables were included in this analysis: (1) relationship to child (mother, father, other); Education level (less than high school, high school/GED, more than high school); (3) Language spoken at home (English, Spanish, Creole, Other); and place of birth (United States or outside of the United States).

Early Childhood Obesity Risk and Protective Measures.

Parents were asked on the baseline intake form about early childhood feeding patterns and child birthweight. Specifically, the were asked (1) Did you breastfeed your child (Y/N)?; (2) If so, how many months did you breastfeed? (range of 0 to 36); (3) Did you bottlefeed? (Y/N); (4) If so, for how many months? (range of 0 to 60); (5) How old was your child when they started eating solid foods? (< 1 month, 2 months, 3 months, 4 months, 5 months, > 6 months); and (5) how much did your child weigh when they were born? (pounds). All measures were available in English, Spanish or Creole and the parent determined what language the interview was delivered in.

Child Body Mass Index.

Assessment of child body composition included height (using a stadiometer) and weight (using a digital scale), which were converted to BMI (weight in kilograms/height in meters squared) and then to a BMI age-and sex-adjusted percentile (Centers for Disease Control and Prevention, 2002). Children were asked to remove shoes and any heavy outer clothing prior to measurement. Weight and height measurements were collected by trained staff and were based on US Department of Health and Human Services guidelines for accurate anthropomorphic measurement (Centers for Disease Control and Prevention, 2002).

Statistical Analysis

For all analyses, we utilized multilevel models, including a random effect for clusters created by children within centers. Thus, defining centers as a random effect also accounts for the trend effect of BMI since children within center were recruited at similar times. To examine ethnic group differences in early childhood behavior, birthweight and BMI, frequencies and means were calculated for each group. Multilevel models examined whether differences in these factors were not statistically significant across both groups. Model estimated means and standard errors and significant between individual group differences are presented. In order to maintain precision, BMI percentile was first analyzed as a continuous variable and then as a categorical variable with three levels, and finally as a binary variable. The BMI percentile was categorized into the following three groups: normal (BMI < 85^tth percentile for age and sex), overweight (BMI ≥ 85^th percentile for age and sex) and obese (BMI ≥ 95^th percentile for age and sex) (Centers for Disease Control and Prevention, 2002). The two level BMI percentile was categorized as follows: normal (BMI < 85^tth percentile for age and sex), and overweight or obese (BMI≥ 85^th percentile for age and sex). Child age was analyzed as a continuous variable, and ethnicity was analyzed as a categorical variable with three levels; African American, Haitian and Other. African Americans were the reference level/group in all models. Sex as a biological variable was analyzed as a binary variable with female as reference. Random effect proportional odd models were fitted to compares these categories. Data were analyzed using SAS 9.4 (SAS Institute Inc., Cary, NC). Statistical tests resulting in a probability less than 0.05 were considered statistically significant.

Results

A total of 370 children from 35 centers (not all centers had children/families who identified as one of our target ethnicities) were included in this analysis. Over half (n=209, 56.5%) self-identified as non-Hispanic black or African American (AA), 120 (32.4%) as Haitian, and 41(11.1%) as Caribbean Islander/West Indian or multiracial. As shown in Table 1 the sample had slightly more females (52.4%) and Haitians and Caribbean Islander/West Indian or multiracial were significantly older than AAs by about 6 months and 7.2 months, respectively. Mean BMI percentile was very similar for Haitian, Caribbean Islander/West Indian or multiracial, and AA children (60.8^th percentile, 62.8^th percentile, and 60.1 percentile, respectively).

Table 1.

Child and Caregiver Demographic and Anthropometric Characteristics by Ethnic Group, Healthy Caregivers, Healthy Children, Phase 1 and 2, 2012–2018.

Characteristic	African American		Haitian		Othera		P
CHILD
	N=209	Mean (SD)	N=120	Mean (SD)	N=41
Age	209	3.1 (1.1)	120	3.6 (1.0)	41	3.7(1.1)	<0.001
Sex	N	Mean (%)	N	Mean (%)
Female	109	47.9	65	45.8	20	51.2	0.83
Male	100	52.1	55	54.2	21	48.8
BMI Percentile	197	60.1(28.0)	117	60.8(30.8)	41	62.8(28.1)	0.87
BMI Percentile Groupb
Normal Weight	162	77.5	84	70.0	30	73.2	0.24
Overweight	28	13.4	19	15.8	9	21.9
Obese	19	9.1	17	14.2	2	4.9
CAREGIVER
Relationship to Child
Mother	179	85.7	95	79.8	39	95.1	0.03
Father	17	8.1	19	16.0	0	0.0
Other	13	6.2	5	4.2	2	4.9
Education Level
< 12th grade education	21	10.1	36	31.9	7	18.0	<0.001
Completed high school/GED	104	50.2	29	25.6	13	33.3
>12th grade education	82	39.6	48	42.5	19	48.7
Language spoken at home
English	186	90.3	19	16.0	26	65.0	<0.001
Other	20	9.7	100	84.0	14	65.0
Birthplace
Foreign born	23	11.2	97	80.8	18	45.0	<0.001
US born	183	88.8	23	19.2	22	55.0

^aCaribbean Islander/West Indian, African, and Multiracial

^bNormal weight = BMI < 85^th percentile, overweight= ≥ 85^th percentile but < 95^thpercentile, obese= ≥ 95^th percentile)

Significant group differences were found for caregiver relationship to child, education level, language spoken at home, and birthplace of caregiver. A subgroup analysis based on independent t-tests revealed no significant differences between Haitian and ‘Other’ group (p-value=0.89) but significant differences in these two groups versus AA (p<0.001). A subgroup analysis to compare observed versus expected caregiver education level based on chi-square revealed significant differences between Haitian and AA (p<0.01) but not between Haitian and ‘Other’ (p=0.24) or AA and ‘Other’ (p-value=0.11). A further sub-group chi-square test for caregiver –child relationship revealed significant differences between Haitian and ‘Other’ (p=0.02) but neither between AA and Haitian (p=0.07) nor AA and ‘Other’ (p=0.15). A chi-square test for observed versus expected value for language spoken at home revealed significant differences across the three groups (p<0.01). Similar differences were observed for birthplace of caregiver (<0.01).

As shown in Table 2, no significant differences in mean BMI percentile between Haitian, Caribbean Islander/West Indian or multiracial and AA preschoolers were found. This might be due to the fact that the study was not well powered as shown by the large variability, and thus standard error of the estimates. Further analyses comparing the three weight categories: normal (body mass index or BMI < 85^th percentile for age and sex), overweight (85^th percentile ≥ BMI < 95^th percentile) and obese (BMI ≥ 95^th percentile) reveal no significant differences. However when gender was taken into account, Haitian and AA male preschoolers were less likely to be overweight or obese as compare to female Haitian and AA preschoolers. A further analysis combining the overweight and obese groups showed similar results. In addition to gender differences, there was a significant age difference; as age increased healthy weight decreased.

Table 2.

Multilevel models comparing (1) Continuous BMI percentile^a; (2) Three BMI levels (normal, overweight and obese); and (3) Binary (normal versus overweight/obese) in NHB subgroups^b controlling for gender and age.

Parameter	Estimate (S.E)b
Model 1 (Outcome Continuous BMI percentilea)
Paremeters
Haitian versus African American	−0.95 (3.74)
Otherc versus African American	0.77(5.11)
Male versus Female	−2.89 (3.07)
Age	2.58(1.62)
Model 2 (Outcome three level BMI percentilea)
Haitian versus African American
	0.05(0.05)
Other versus African American	−0.01(0.07)
Male versus Female	0.09(0.04)*
Age	0.04(0.02)*
Model 3 (Outcome Binary BMI percentilea)
Haitian versus African American	−0.05(0.05)
Other versus African American	−0.02(0.08)
Male versus Female	0.10(0.04)*
Age	−0.04(0.02)*

^aNormal BMI < 85^th weight = percentile, overweight= ≥ 85^th percentile but < 95^thpercentile, obese= ≥ 95^th percentile)

^bBoldface indicates statistical significance (*p<0.05)

^cCaribbean Islander, West Indian, African, More than One Race or Non-Hispanic Black Ethnic Group

Table 3 summarizes multilevel models comparing child BMI percentile in the 3 subgroups controlling for several caregiver characteristics (education level, mother versus father-led care, English versus other as first language, and foreign or US-born caregiver). The only statistically significant finding was that US-born parents had children with a BMI almost 9 percentile points less than foreign-born parents. Age was borderline significant with p-value=0.06.

Table 3.

Multilevel models comparing BMI percentile in Non-Hispanic Black (NHB) subgroups.^a

Parameter	Estimate (SE)
Model (Continuous BME percentile as outcome)
Haitian versus African American	−3.96(5.34)
Otherb Caribbean versus African American	−1.68(5.50)
High school/GED versus < 12th grade	0.40 (4.58)
>High school versus <12th grade education	5.19(4.41)
Father versus Mother providing care	−2.03 (5.30)
Other versus Mother providing care	−7.16 (6.85)
English as first language versus Other	4.14 (4.90)
Caregiver born in USA versus Foreign born	−9.13 (4.61)*
Male versus Female	−2.42(3.14)
Age	3.08(1.64)

^*statistical significance (*p<0.05)

^accontrolling for caregiver relationship to child, education level, language spoken at home and birthplace of caregiver.

^bCaribbean Islander, West Indian, African, Multiracial

Table 4 presents multilevel models comparing parental feeding behaviors and birthweight in the 3 NHB subgroups controlling for caregiver education, caregiver relationship to child, language spoken at home and birthplace of caregiver. Haitians and ‘Other’ were more likely to breastfeed compared to AA (p<.001). There was no significant difference between Haitian and other. No other significant group differences were found for all behaviors analyzed.

Table 4.

Multilevel Models Comparing Early Childhood Feeding Behaviors and Birthweight in Non-Hispanic Black (NHB) subgroups.^a

Characteristic	African American		Haitian		Otherb		p
	N	Mean (%)	N	Mean (%)	N
Breastfed							<0.01
No	98	52.4	39	35.1	14	37.8
Yes	89	47.6	72	64.9	23	62.2
Breastfeeding duration, months	209	7.1(6.3)	120	7.7(6.2)	41	5.5(3.6)	0.41
Bottle feeding duration, months	209	8.6(3.8)	120	7.9(4.3)	41	7.9(3.5)	0.59
Age solid foods introduced, months	209	7.9(3.0)	120	6.7(3.1)	41	7.0(2.1)	0.06
Birthweight, pounds	209	6.3(2.3)	120	6.8(1.6)	41	6.6(1.1)	0.08

^acontrolling for caregiver education, caregiver relationship to child, language spoken at home and birthplace of caregiver.

^bCaribbean Islander, West Indian, African, Multiracial

Discussion

The present analyses explored the relationship between parent characteristics, early childhood obesity-related risk and protective behaviors and child BMI percentile among preschool age children from various non-Hispanic black subgroups, and primarily originating from the Caribbean region. Overall, pediatric obesity health disparities continue to impact the U.S., and especially the NHB population. However, the label “NHB” refers to a number of different groups that may be quite distinct from one another in terms of feeding practices (among other variables). We found no differences in mean child BMI percentile for the 3 subgroups, but children of US-born parents had a BMI almost 9 percentile points less than foreign-born parents despite all three groups being born at very similar birthweights. Haitian women were significantly more likely to breastfeed versus other NHB and AA women.

Despite no significant differences in the prevalence estimates for BMI percentile groups between non-Hispanic black subgroups of preschoolers, our results suggest the need to continue examining the complex early feeding behavioral diversity between these groups. Although previously literature shows ethnic group differences in risk and protective-related child feeding behaviors between NHB, Non-Hispanic White and Hispanic groups, this study suggests that early childhood risk and protective factors for obesity may be similar, with the exception of breastfeeding practices, across NHB subgroups.

Surveillance estimates of US breastfeeding initiation and duration have consistenly shown that NHB infants breastfeed less than other racial/ethnic groups (Beauregard, 2019). As found in previous work, Haitian women were more likely to initiate breastfeeding and have a longer duration of breastfeeding versus their NHB group counterparts (Messiah et al. 2015). Previous studies have shown that Haitian women acknowledge breast milk as the best initial food for their infants (Thomas & DeSantis, 1997; Jenny & Kelly, 2013). However, Haitian women report job-and mother-related, cultural-belief barriers to breastfeeding such as working outside the home, fear of transferring mother’s sicknesses to infant, and lack of desire or patience. Due to these cultural beliefs, Haitian Americans may be likely to substitute whole milk, and other supplemental liquids and foods, in place of infant formula and supplemental liquids and foods (Jenny & Kelly, 2013).

Previous studies have identified earlier introduction of solid food as a risk factor for obesity (Taveras et al. 2013). A possible rationale for the absence of findings for Haitians may include differences in how cultural identity, values, and practices influence early childhood feeding behaviors in comparison to African Americans. For instance, it has been established that Haitian Americans consume significantly higher rates of fruits and vegetables than AAs, but Haitian American diets also contain higher carbohydrates, saturated fat, sodium, and calories from solid fats and added sugars (Huffman et al., 2011).

Although a dearth of literature in NHB subgroups is apparent, some researchers have begun to examine the differences in health profiles between African immigrants and African Americans. One recent study using data from the National Health Interview Surveys found that African immigrants had lower age‐standardized hypertension, diabetes mellitus, overweight/obesity, high cholesterol, and current smoking prevalence than African Americans (Turkson‐Ocran et al., 2020). Another study found that foreign-born blacks experience lower odds of obesity but higher odds of diabetes than US-born blacks in New York City (Horlyck-Romanovsky et al., 2019). According to Brown (2015), about ten percent of the self-identified NHB population in the US were born outside of the country; however, that estimate is projected to nearly double by 2060. In fact, NHB immigrants make up a considerable proportion of the overall NHB population in large metropolitan areas. For instance, NHB immigrants comprise 28% of the NHB population in New York and the South Florida metropolitan area houses the largest enclave of Haitians living in the U.S., accounting for nearly a third of the region’s NHB population.

Furthermore, the US population continues to become ethnically diverse, and research agendas must reflect the current racial/ethnic composition of the nation. Distinct diasporic populations must have access to culturally attuned strategies, programs, and resources which help deal with health issues. Limited research on obesity prevention has been conducted with NHB subgroup families, and those originating from the Caribbean Basin in particular and even fewer studies have compared NHB subgroups to one another. Our results may therefore be informative regarding specific early-childhood obesity prevention intervention strategies for African Americans versus Haitian Americans. Given approximately 254,600 NHB children attend preschool in the US (Barnett et al., 2014), this setting may provide families with an opportunity to receive interventions aimed at promoting healthy child feeding behaviors.

Study Limitations

The present results must be interpreted in light of a number of important limitations. Because this was a cross-sectional analysis, there can be no inferences of causality. Furthermore the sample size of the ‘Other’ group was small thus limiting precision and generalizability of findings. Future studies might use larger samples, natural-history longitudinal designs, and a wider variety of NHB subgroups (e.g., Jamaican, Nigerian, Somalian) in addition to African Americans and Haitians to increase generalizability. Future studies may also consider adding more contextual variables, such as those that focus on acculturation and early childhood obesity risk and protective factors and behaviors from the caregiver’s native country.

Despite these and other limitations, our study is an innovative effort to disaggregate NHB groups to examine feeding practices and their effects on young child BMI. Many times these smaller groups are simply excluded from analyses entirely yet deserve to be included given the ever increasing diversity of our nation. Our findings may help inform culturally tailored interventions to promote the maintenance of healthy weight in early childhood, thereby potentially reducing obesity and related disease rates later in development, and life.

Conclusions

Despite no significant differences in the prevalence estimates for BMI percentile groups between non-Hispanic black subgroups of preschoolers, our results suggest the need to continue examining the complex early feeding behavioral diversity between these groups given the significant proportion of unhealthy weight children in all groups. Findings here suggest that NHB race classification can identify important subgroup behavioral similarities which in turn may inform culturally-tailored strategies to promote healthy weight in early childhood. Foreign-born caregivers in particular may benefit from healthy weight promotion information, and as early as possible in their child’s life and development.

Significance:

Obesity studies in preschool-age children have included non-Hispanic Black (NHB) children but few have investigated between-subgroup differences even though there may be cultural risk and protective practice differences, challenging the generalizability of findings. Our findings reported here suggest that NHB race classification can identify important subgroup behavioral similarities as well as the variations in how risk and protective factors operate, which in turn may inform culturally-tailored strategies to promote healthy weight in early childhood. Foreign-born caregivers in particular may benefit from healthy weight promotion information, and as early as possible in their child’s life and development.