Design  Cross-sectional analysis of data collected by the US Department of Education, National Center for Education Statistics.

Setting  A nationally representative sample of infants enrolled in the Early Childhood Longitudinal Study, Birth Cohort at baseline.

Participants  A total of 8150 infants aged 9 months.

Main Exposures  Age (in months) at initiation and type and intensity of child care.

Outcome Measures  Breastfeeding initiation, early introduction of solid foods (<4 months), and weight gain (birth to 9 months).

Results  A total of 55.3% of infants received regular, nonparental child care and half of these infants were in full-time child care. Among infants in child care, 40.3% began at younger than 3 months, 39.3% began between 3 and 5.9 months of age, and 20.7% began at 6 months or older. Infants who initiated child care at younger than 3 months were less likely to have been breastfed (odds ratio, 0.58; 95% confidence interval [CI], 0.43-0.74) and were more likely to have received early introduction of solid foods (odds ratio, 1.73; 95% CI, 1.43-2.04) than those in parental care. Infants in part-time child care gained 175 g (95% CI, 100-250 g) more weight during 9 months than those in parental care. Infants being cared for by relatives had a lower rate of breastfeeding initiation, a higher rate of early introduction of solid foods, and greater weight gain compared with infants receiving parental care. The early introduction of solid foods was a risk factor for weight gain.

Conclusions  Child care factors were associated with unfavorable infant feeding practices and more weight gain during the first year of life in a nationally representative cohort. The effects of early child care on breastfeeding and introduction of solid foods warrant longer follow-up to determine subsequent risk of childhood overweight.

Child care experience has been associated with cognitive, language, and socioemotional development and behaviors during childhood.^6-7 Several randomized studies have shown the positive effects of nonparental child care on cognitive development and academic success, even though the optimal timing of child care initiation remains unclear.⁸ Previous literature on the effects of child care on physical health has largely been confined to infectious diseases. Children in day-care centers are at a higher risk of having several adverse health outcomes, such as respiratory tract illness, ear infection, and enteric illness, than children cared for in homes.^9-10 Despite the increased use of early child care, to our knowledge, no information exists about the potential influences of earlier initiation, type, or intensity of child care on infant feeding practices and development of overweight. To our knowledge, only 1 study¹¹ has examined the relationship between child care attendance during the preschool period and later risk of overweight among US children aged 6 to 12 years, finding that limited center-based child care attendance (<15 h/wk) was associated with a lower risk for subsequent overweight. The possible effects of infant exposure to child care on feeding practices and the development of overweight remain to be investigated.

Both breastfeeding and early introduction of solid foods are important infant feeding practices that influence growth and the development of overweight. Accumulating epidemiologic studies^12-13 show that breastfeeding could lower the risk of overweight during childhood. A few studies have suggested that early introduction of solid foods is associated with greater risk of overweight. Wilson and colleagues¹⁴ reported increased body weight and obesity at the age of 7 years among children who had received solid foods before the age of 15 weeks, despite no apparent effect on weight at the age of 2 years. A prospective study¹⁵ among 3768 mother-infant dyads from the Danish National Birth Cohort found that both shorter breastfeeding duration and earlier introduction of solid foods (at an age of <16 weeks) were associated with increased infant weight gain.

As today's infants and toddlers spend more hours in nonparental child care, their food consumption consequently is influenced by the foods that child care providers offer. According to the Feeding Infants and Toddlers Study,¹⁶ total energy intake at lunch was highest for those at day care (332 kcal) compared with those who ate away from home (308 kcal) or ate at home (281 kcal). This study also indicates that many US infants and toddlers already consume foods outside the home and the composition of food intake needs to be improved. Food-related experiences in the first 2 years of life, such as longer breastfeeding and variety of fruit exposure, also predict more variety of fruit intake among school-aged children.¹⁷ Considering that food preferences and eating habits begin to be formed during infancy, both parents and child care providers can play an important role in fostering children's healthy eating habits. To our knowledge, the relationship among child care factors and infant feeding practices and risk of overweight has not been examined.

The aim of this study is to consider the effects of infant child care characteristics (age at initiation, types, and intensity) on breastfeeding, early introduction of solid foods, and weight gain by using a nationally representative sample of infants aged 9 months enrolled in the Early Childhood Longitudinal Study, Birth Cohort (ECLS-B). We hypothesized that early nonparental child care could decrease breastfeeding rate and/or lead to earlier introduction of complementary solids and that infants in child care may have a greater rate of weight gain than those with parental care.

METHODS

The analyses were performed using baseline data (9-month assessment) from the ECLS-B conducted by the US Department of Education, National Center for Education Statistics.¹⁸ The study population is a representative sample of infants born in 2001 in the United States, with an oversampling of Asian and Pacific Islanders, American Indians, twins, and infants with moderately low and very low birth weight (LBW). The sample of infants was selected using a clustered, list frame sampling design. The list frame comprised births registered in the National Center for Health Statistics' vital statistics system from lists provided by state registrars. Infants whose birth mothers were younger than 15 years at the child's birth were excluded. Among infants who participated in the 9-month assessment, we selected infants 8 to 14 months of age (n = 9650) and excluded premature infants who were born at less than 32 weeks' gestation (10% of the sample). After further exclusion of infants without measures of weight, length, and birth weight (3%) and those with biologically implausible values for weight, length, and weight for length (3%),¹⁹ the final analytic sample was 8150. All numbers presented in this study were rounded to multiples of 50 to comply with the ECLS-B data security requirement. Our study population comprises infants of gestational age older than 32 weeks who were born in the United States in 2001 to mothers older than 15 years and who were not adopted or deceased by the age of 9 months.

DATA COLLECTION

The ECLS-B research team collected baseline data from October 2001 through December 2002 during a visit to the child's home when the child was approximately 9 months old. Trained interviewers measured each infant's length and weight and administered the survey questionnaire as part of the parent interview. A measure mat (a pediatric length-measuring device) and a digital bathroom scale were used to obtain the length and weight of the child. The child's weight was determined by weighing the mother while she was holding the child and then subtracting the mother's weight. The primary caregiver (99% were the biological mother) provided information on whether the child was in child care, the type of child care, the number of hours in child care, and the age at which the child first entered care.

The ECLS-B defines child care as "infants receiving care on a regular basis from persons other than their parents." Child care includes regular care and early childhood programs, whether or not there is a charge or fee, but not occasional babysitting. We examined the 3 child care arrangements as main exposure variables, with parental care as the reference group. Age at initiation of child care indicates the earliest age the child first began any type of nonparental care on a regular basis (0-2.9, 3.0-5.9, or 6.0 months); "parental care" is a reference group of those who do not send their infants to a child care facility. The rationale for this breakdown is because maternity leave of 3 months is common. We further divided the period into 3.0 to 5.9 months and 6.0 months or more to provide an equal distribution. Child care type indicates the primary, nonparental, individual child care arrangement in which the child spent the most hours per week at the time of the 9-month home visit (parental care, relative care, nonrelative care, or center-based care). Child care intensity indicates the total number of hours per week the child spent in all primary and secondary care arrangements (parental care, part-time care [<35 h/wk], or full-time care [35 h/wk]). The primary outcome variable was weight gain between birth and 9 months of age. The prevalence of overweight (weight for length 95th or <95th percentile referenced from the Centers for Disease Control and Prevention 2000 growth charts)²⁰ was also estimated.

Secondary outcome variables were breastfeeding and early introduction of solid foods. Breastfeeding questions ask whether breastfeeding was ever initiated (yes or no) and when the first time formula or cow's milk was introduced (child's age in months). In the questionnaire, solid foods referred to cereal and baby food in jars but not finger foods. Mothers indicated whether one of the solid foods was introduced to the infant and at what month of the child's age the food was introduced. We defined early introduction of solids as one of the solid foods being introduced to the infant at earlier than 4 months of age. A cutoff point of 4 months for early introduction of solid foods is in accordance with the American Academy of Pediatrics' recommendation that solid foods be introduced between 4 and 6 months of age.²¹ Maternal employment status was derived from 3 questions in the baseline parental interview: (1) During the past week, did you work at a job or business for pay? (2) How many jobs do you have now? (3) About how many total hours per week do you usually work for pay (counting all jobs)? Our maternal employment variable was current (at child's age of 9 months) work status, quantifying the number of hours per week the mother worked (part-time [<35 h/wk] or full-time [35 h/wk]).

MEASURES AND ANALYSIS

Maternal weight status and birth weight have been suggested as 2 strong indicators of a child's risk of overweight. The maternal prepregnancy body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) is associated with a child's weight gain between birth and 1 year of age¹⁵ and later childhood.^22-23 The birth weight and risk of overweight have been shown to have a U-shaped relationship, with a higher prevalence of obesity for the lowest and highest birth weights.^{22, 24-25} We treated maternal BMI and birth weight as potential confounders of our proposed hypotheses and adjusted for them in the multivariate regression analyses. To address the differential weight gains by birth weight status, we stratified analyses of the main models among the subgroups of infants who were born before 37 weeks' gestation and at less than 2500 g of birth weight (preterm and LBW, 3.5% of infants) and those born at 37 weeks' gestation or later and at 2500 g or greater of birth weight (term and normal birth weight [NBW], 87.6% of infants). Other confounders included in the models were maternal smoking status (ever smoked or not), maternal education (less than high school, high school graduate, some college, college or graduate school graduate), parental marital status (married or not married), household poverty (<100% or 100% of the poverty threshold),²⁶ child's age (in months), sex, race/ethnicity (white, black, Hispanic, Asian, and other), prematurity (<37 vs 37 weeks), and birth weight (in grams).

The full-sample weight was used for all estimates to represent the US study population. For the final multivariate regression models, we used the Jackknife 2 standard error estimation to account for multistage sampling and unequally weighted designs using AM software (version 0.06.0.3.beta, June 17, 2005, American Institute for Research and Jon Cohen, National Center for Education Statistics of US Department of Education). Other analyses were performed with the SAS statistical analysis program, version 9.1 (SAS Institute Inc, Cary, North Carolina). The study protocol was reviewed and exempted by institutional review board of the Harvard School of Public Health.

RESULTS

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 1. Demographic Characteristics, Infant Feeding Practices, and Weight Gain by Child Care Factors Among US Infants Aged 9 Monthsa

The effects of child care factors on infant feeding practices and weight gain after controlling for potential confounders are given in Table 2. All child care factors were significantly associated with infant feeding practices and weight gain. Infants initiating child care before 3 months of age had 1.73 times (95% confidence interval [CI], 1.43-2.04 times) higher odds of having early introduction of solid foods than those in parental care. Also, they were less likely to have breastfeeding initiation (odds ratio, 0.58; 95% CI, 0.43-0.74). Infants in parental care were more likely to have breastfeeding initiated and solid foods introduced after 4 months of age compared with those in child care settings. Infants in part-time child care gained 175 g (95% CI, 100-250 g) more weight in the first 9 months than those in parental care. Infants receiving care from relatives showed more weight gain (162 g; 95% CI, 106-219 g) compared with those in parental care. Only infants who initiated child care from 3 to 6 months of age gained more weight (155 g; 95% CI, 87-225 g) compared with those in parental care. Because birth weight and prematurity may modify the observed relationship, we conducted subgroup analyses among infants who were preterm and had an LBW and those who were term and had an NBW. Preterm and LBW infants gained more weight if they started child care before 3 months of age, were in center-based care, and were in full-time care compared with those in parental care. Term and NBW infants gained more weight if they started child care before 9 months of age, were in nonrelative and relative care, and were in part-time care compared with those in parental care.

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 2. Effects of Child Care Factors on Infant Feeding Practices and Weight Gain Among US Infants Aged 9 Months

We also examined whether infant feeding practices predicted weight gain (Table 3). We examined breastfeeding as a binary variable (breastfeeding initiation or not) and examined the continuous variable of breastfeeding duration (in months). Because breastfeeding initiation was significantly related to child care factors in the univariate models, we used breastfeeding initiation in the remaining analyses. We found no interaction between breastfeeding initiation and early introduction of solid foods in predicting weight gain. Findings were consistent with lack of association of breastfeeding initiation to weight gain among all infants and term and NBW infants, whereas breastfeeding initiation was associated with less weight gain among preterm and LBW infants. The association of early introduction of solid foods and weight gain was significant among all infants. Because the early introduction of solid foods was related to both child care factors and weight gain, we evaluated this variable as a potential mediator and not a confounder on the relationship between child care arrangements and weight gain. We conceived of child care arrangement as influencing feeding practices in a temporal manner and, hence, it is part of the causal pathway to an increase in weight gain. However, we found no mediating role for early introduction of solid foods in the hierarchical regression models of weight gain (data not shown).

View this table: [in this window] [in a new window] [as a PowerPoint slide]   Table 3. Effects of Infant Feeding Practices on Weight Gain Among US Infants Aged 9 Months

We estimated the expected weight gains for the highest and the lowest weight gain of a hypothetical 9-month-old infant based on our predicted model. We fixed the child's birth weight at 3500 g, the age at 9 months, and the mother's BMI at 25. The highest weight gain of a 9-month-old infant would be 6798 g for a preterm Hispanic boy who was not breastfed, had solid food introduced before 4 months of age, and had child care initiated between 3 and 6 months of age by a married, smoking mother who had graduated from high school, was working part-time, and was from a family that was not poor. The lowest weight gain of a hypothetical 9-month-old infant would be 5172 g for a term Asian girl in whom breastfeeding had been initiated, in whom solid foods had been introduced after 4 months of age, who was in parental care, and who was born to a single, nonsmoking mother with some college education who was working full-time and was from a poor family.

COMMENT

Overwhelming and consistent data support the notion that early weight gain during infancy is a strong risk factor for overweight in childhood and adulthood.^27-29 In particular, rapid weight gain may be especially harmful to children born with an LBW. Early accelerated weight gain among children with an LBW has been associated with a higher BMI, especially central obesity in early childhood.³⁰ Another longitudinal study³¹ showed that children who were above the 50th BMI percentile at the age of 2 to 5 years continued to increase in BMI percentile and were more likely to become overweight at the age of 12 years. Overweight during childhood is associated with comorbidities, such as asthma, hyperlipidemia, hypertension, and type 2 diabetes mellitus,^32-33 as well as with higher adulthood morbidity and mortality.³⁴ Previous research shows that we need to consider weight status as a continuous spectrum originating early in life and suggests that the infancy and preschool years may constitute a key period for the development of overweight and its long-term health consequences. Thus, it is important to avoid excessive weight gain for both ends of birth weight distribution during infancy. However, studies on how to prevent overweight in infancy are scarce and, more important, do not take into account the reality that infants are increasingly being taken care of outside the home, in child care settings.

All child care factors were consistently related to infant feeding practices after controlling for other confounding factors. We found no protective effect of breastfeeding on weight gain, in contrast to findings in other studies,³⁵ but we did find a coherent association between early introduction of solid foods and weight gain, similar to other published research.^{15, 35} In our subgroup analysis, we found the protective effect of breastfeeding but no effect of introduction of solid foods on weight gain among preterm and LBW infants. We were not able to show the potential mediating role of infant feeding practices between the observed risk of child care and weight gain. Our cross-sectional study design limits the detection of a potential mediating role of feeding practices on weight gain in the child care setting. In addition, we do not know the frequency, amount, and type of solid foods infants consumed. Other child care factors, such as quality or nutrition and physical activity policies, might play an important role in feeding practices and risk of overweight, but we do not have information in the ECLS-B to address this question.

Although the effect sizes and long-term effects of child care quality on developmental outcomes are debatable, studies^36-37 consistently have shown that children receiving high-quality care performed better in social and intellectual competencies, whereas children receiving low-quality care fared worse, compared with children receiving parental care. To our knowledge, only 16 states have or are in the process of developing child care credentials for infant and toddler age groups according to the 2005-2007 Childcare and Development Fund plans.³⁸ A study³⁹ of child care centers in 4 states found that 40% of centers serving infants and toddlers were rated as poor. The Study of Children in Family Childcare and Relative Care⁴⁰ found that only 9% of family child care homes in 3 states were rated as good quality and 35% of the homes were rated as inadequate, raising a concern about the early child care arrangement for infants and toddlers. Parents were not well aware of the child care their children received and usually overrated its quality, according to the Cost, Quality, and Child Outcome Study.⁴¹ In short, although child care quality matters tremendously in child development, we do not know how important it may be in infant nutrition and growth outcomes. Nevertheless, the quality of child care needs to be improved to ensure a healthy eating and growing environment.

Recent Start Healthy Feeding Guidelines provide recommended infant and toddler feeding practices for parents and caregivers, such as when, what, and how to feed complementary foods.⁴² One of the first studies to examine the child care environment as a context for promoting healthy weight, the Nutrition and Physical Activity Self-assessment for Child Care pilot intervention, was developed to improve nutrition and the physical activity environment in center-based care for preschool-aged children.^43-44 However, infant feeding practices in child care settings are largely unknown. Our study results provided new evidence that child care influences both infant feeding practices and risk of overweight at least during infancy. Thus, more research is needed to understand the mechanisms by which these early child care factors and infant feeding practices affect subsequent risk for childhood overweight.

AUTHOR INFORMATION

Accepted for Publication: November 16, 2007.

Author Contributions: Study concept and design: Kim and Peterson. Acquisition of data: Kim. Analysis and interpretation of data: Kim and Peterson. Drafting of the manuscript: Kim. Critical revision of the manuscript for important intellectual content: Peterson. Statistical analysis: Kim. Obtained funding: Kim and Peterson. Administrative, technical, and material support: Kim and Peterson. Study supervision: Kim.

Funding/Support: This study was supported in part by the Berkowitz Fellowship of the Department of Nutrition, Harvard School of Public Health, an ECLS-B cohort training grant, the National Center for Education Statistics, and Training Grants on Statistical Analysis for Education Policy from the American Educational Research Association.

Role of the Sponsor: The funding bodies had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: Jennifer Park, PhD, US Department of Education, and Amy Rathbun, PhD, American Institutes for Research, provided assistance related to the sample design, data collection process, and measures included in the ECLS-B; and Jon Cohen, American Institutes for Research, provided technical assistance.