|
 |
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2008 American Public Health Association
DOI: 10.2105/AJPH.2006.101683
RESEARCH AND PRACTICE |
Sandra B. Procter is with Kansas State University Research and Extension, Department of Human Nutrition, Kansas State University, Manhattan. Carol Ann Holcomb is with the Department of Human Nutrition and the interdepartmental Master of Public Health program, Kansas State University, Manhattan.
Correspondence: Requests for reprints should be sent to Sandra B. Procter, PhD, RD, K-State Research and Extension, Department of Human Nutrition, 204 Justin Hall, Manhattan, KS 66506 (e-mail: procter{at}humec.ksu.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Objectives. We sought to determine whether increased duration of breast-feeding was associated with decreased risk of overweight among 4-year-old children in Kansas families with limited means.
Methods. We linked data on Kansas families from the Pediatric Nutrition Surveillance System and Pregnancy Nutrition Surveillance System, 1998 to 2002, to determine breastfeeding duration and weight status at age 4 years. Overweight among 4-year-old children was defined as body mass index–for-age at the 95th percentile and above. Linked analysis included 3692 children. Logistic regression was applied to determine odds of being overweight at each duration of breastfeeding.
Results. Breastfeeding duration considered independently showed a significant protective association with childhood overweight at age 4 years for all non-Hispanics (odds ratio [OR]=0.72; 95% confidence interval [CI]=0.55, 0.94) and for Whites only (OR=0.68; CI=0.50, 0.92). When we controlled for other significant risk factors for childhood overweight, the association diminished and was not statistically significant.
Conclusions. Although breastfeeding for longer duration appeared to be protective against overweight among 4-year-old non-Hispanic children, cultural and environmental factors may override this protective benefit.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Childhood overweight may be the most significant health issue facing American children today. The number of children who are overweight in the United States has risen dramatically since the 1970s, and the prevalence of excess weight in US children continues to increase.1 Excess weight has negative effects on children from early childhood throughout life; therefore, research regarding causes as well as prevention and treatment of childhood overweight is needed.
Childhood overweight reflects the convergence of many biological, economic, and social factors. Each of these factors affects family food choices, food preparation, and food consumption, and together they contribute to the complex problem of overweight in children. No single intervention is likely to produce large reductions in the prevalence of overweight children,2 but choosing to breastfeed an infant rather than use formula is one factor that may prevent subsequent overweight.3 Recent studies have reported that longer breastfeeding is protective against childhood overweight.4–9
In 2004, Howell Wechsler, acting director of the Centers for Disease Control and Prevention Division of Adolescent and School Health, testified before a congressional subcommittee that the "promotion of breastfeeding and efforts to increase its duration" is one behavioral strategy for reducing rates of overweight and obesity in children and adolescents.10 Data from 2002 revealed that breastfeeding rates among recipients of benefits from the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) lagged behind those of non-WIC recipients by more than 20% in the hospital and at 6 months of age.11 Kansas data showed similar disparity between in-hospital breastfeeding rates (initiation) and breastfeeding rates at 6 months (duration) for WIC infants, with only 19.6% still breastfeeding at 6 months, below even the national rate for WIC infants of 22.1%.12 Breastfeeding is initiated in this population, but it is not continued.
Grummer-Strawn and Mei reported that prolonged breastfeeding was associated with a reduced risk of overweight in non-Hispanic White children.4 Their study analyzed linked data from the Pediatric Nutrition Surveillance System (PedNSS) and the Pregnancy Nutrition Surveillance System (PNSS) for 7 states. The PedNSS and PNSS receive data from publicly funded health and nutrition programs, including WIC.
Little was known about breastfeeding and childhood overweight among Kansans, particularly low-income Kansans. Kansas was not included in the 2004 study by Grummer-Strawn and Mei.4 Therefore, we designed our study to address the paucity of data for Kansas by analyzing linked data from the PedNSS and PNSS to determine whether longer breastfeeding was associated with decreased incidence of childhood overweight at the age of 4 years among a Kansas WIC population.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Data Sets
Data from the Kansas PedNSS master file for years 1998 through 2002 and the Kansas PNSS master file for children born in 1998 were obtained from the Centers for Disease Control and Prevention.
The PedNSS is a public health surveillance system that describes the nutritional status of low-income US children attending federally funded maternal and child health and nutrition programs.13 The PedNSS contains information on birthweight, breastfeeding (including duration), and race/ethnicity, as well as height and weight data recorded during each visit to public health programs. The data captured information for all children born in 1998 and each of their subsequent visits to the WIC program through 2002. Body mass index (BMI; calculated as weight in kilograms divided by height in meters squared)–for-age was calculated from data obtained on the visit closest to the child’s fourth birthday. Because the PedNSS has multiple records on children in each calendar year, 1 record for each child for each year was selected to avoid double counting.14
The PNSS is a public health surveillance system that monitors risk factors associated with infant mortality and poor birth outcomes among low-income pregnant women who participate in federally funded public health programs.13 The PNSS contains data collected during the mother’s visits, including her date of birth, age at delivery, education level, prepregnancy weight (self-reported), height, prepregnancy BMI, weight gain during pregnancy, alcohol use before and during pregnancy, number of previous pregnancies, infant’s birthweight, and age when solid foods were added to the infant’s diet. The 1998 data captured selected variables for each Kansas WIC participant who delivered an infant that year.
Unlike in many other states in the national WIC data system, in Kansas it is possible to link a child to its mother by matching a unique 5-digit family identification number for the child in the PedNSS that is also embedded in a 7-digit number assigned to the mother in the PNSS. We identified the 1998 birth cohort from the PedNSS, tracked its members through 2002, and linked them to mothers from the 1998 PNSS file.
Statistical Methods and Analysis
To determine which factors were associated with overweight at 4 years of age, characteristics of both mother and child determined from previous research were drawn from the data sets for analysis. For initial comparison with Grummer-Strawn and Mei’s results,4 we kept the same categories for birthweight and breast-feeding duration. Birthweights were categorized as less than 2500, 2500 to 3999, or 4000 g or more. Breastfeeding duration was categorized as less than 4, 4 to 11, 12 to 23, 24 to 51, or 52 weeks or more. For BMI-for-age we were interested in 1 additional category not reported by Grummer-Strawn and Mei: at risk for overweight. This category included children with a BMI-for-age between the 85th and 94th percentiles. Therefore, we divided BMI-for-age into 4 categories: lower than 5th percentile (underweight), 5th to 84th percentile (normal weight), 85th to 94th percentile (at risk for overweight), and 95th percentile or higher (overweight). We based our definitions of these weight categories for 4-year-old children on the 2000 Centers for Disease Control and Prevention gender-specific growth charts.15
We used the linked data from the PNSS to identify and categorize several maternal characteristics.We used the 
2 procedure to analyze differences in characteristics between the 22804 children in the PedNSS data who were born in 1998 and the 3692 children in the linked data set. This analysis contained the following variables: gender, race/ethnicity, birthweight, BMI-for-age, and breastfeeding duration.
We used the 2 analysis to assess differences in the proportion of children in each of the 4 weight categories according to duration of breastfeeding. Mean BMI derived from the raw values for BMI at 4 years of age were compared across all 6 lengths of breastfeeding with a 1-way analyses of variance (repeated-measures analysis of variance) procedure.
Five variables identified in the research literature as risk factors for overweight at 4 years of age were the child’s gender, race/ ethnicity, and birthweight; the mother’s prepregnancy BMI; and formula use versus breastfeeding duration. We included these 5 factors in a univariate analysis of independent association with overweight at 4 years of age in our linked data set. We calculated odds ratios (ORs) with 95% confidence intervals (CIs) using binary logistic regression to determine which factors could be considered as confounding variables.
In the final analysis, breastfeeding duration was categorized into tertiles to provide a relatively even distribution of children at each level. Crude and adjusted ORs and 95% CIs were calculated for 3 samples stratified by ethnicity. Initially, all sample participants were included in both models; Hispanics were then eliminated in both models, and finally, only Whites were included in both models. The P for trend in each of the 6 subgroups was determined with the Wald statistic. All analyses were performed with SPSS for Windows version 13 (SPSS Inc, Chicago, Ill).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Population Characteristics
Table 1
compares the total sample in the PedNSS data set with the subset of participants linked to mother’s data from the PNSS. There was no significant difference in the proportion of male and female infants in the 2 groups. There was a significantly higher proportion of Hispanics in the linked data set (28.1%) than in the original PedNSS data set (21.6%). Although these levels were 3 to 4 times higher than were those found in the Kansas general population (7.0%) in 2000,16 they were reflective of the relatively high number of Hispanics participating in the Kansas WIC program, which reached 30.6% in 2006 (e-mail communication from Patricia Dunavan, nutrition education specialist, Kansas WIC, March 27, 2006).
|
No significant difference was found in the proportion of infants born in each of the 3 birthweight categories in the 2 sample sets. However, BMI-for-age differed significantly between the PedNSS data and the linked data, with children from the linked data set less likely to be underweight (3.5% vs 5.5%) or normal weight (65.5% vs 73.7%) and more likely to be at risk for overweight (15% vs 12.6%) or overweight (16% vs 8.2%).
We found that significantly fewer children were never breastfed in the linked data subset (36.3%) than in the original PedNSS data set (49.5%). In addition, significantly more children were breastfed for 52 weeks or more in the linked data group (11.9%) than in the PedNSS data set (2.4%).
Data Characteristics
Table 2 summarizes the sample distribution by duration of breastfeeding in the linked data sets as well as the distribution across weight categories, from overweight to underweight. Table 2 also shows the mean BMI-for-age by breastfeeding duration. No significant difference was found in the proportion of participants in each weight category by breastfeeding duration. We found no significant difference in the mean BMI-for-age of the 4-year-old children by breastfeeding duration. Increased breastfeeding decreased mean BMI variance, with the BMI variance nearest the mean for those children breastfed the longest. An emergent pattern—the trend of an inverse association between increased breastfeeding duration and decreased proportion of overweight at 4 years of age—is shown but is not consistent across the sample as a whole.
|
Table 3
shows the results of univariate analyses of potential confounding factors or variables other than the main exposure or independent variable of breastfeeding duration that might be associated with overweight at 4 years of age. Three separate analyses with logistic regression were performed. ORs are shown with tests of significance reflected within the 95% CIs for each OR. For each variable, the first category listed was set as the null or reference category. In the total linked sample, all 5 factors (gender, ethnicity, birthweight, mother’s BMI, and formula use) were significantly associated with overweight at 4 years of age. In the second analysis, data for Hispanics were eliminated, and the remaining 4 variables were significant. In the third analysis for Whites only, 3 factors (birthweight, mother’s BMI, and formula use) were found to be significantly associated with overweight at 4 years of age. These variables were considered confounding variables in the subsequent multivariate models.
|
Table 4
presents multivariate analyses of independent variables in the model with the outcome variable of overweight the dependent variable. We performed 3 separate multivariate logistic regression procedures by race/ethnic group. The unadjusted model contained only 2 variables: the main exposure variable of breastfeeding duration at 3 exposure levels and the outcome variable of overweight at 4 years of age. The fully adjusted models contained the significant variables from the univariate analyses (Table 3
).
|
Considered independently, breastfeeding duration showed a significant protective association with overweight at 4 years of age for all non-Hispanics (OR = 0.72; 95% CI = 0.55, 0.94) and for Whites only (OR = 0.68; 95% CI = 0.50, 0.92). When we controlled for other significant risk factors for overweight at 4 years of age, the association diminished and was not statistically significant.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Before we adjusted our model, breastfeeding for 9 weeks or more was found to be mildly protective against overweight at 4 years of age among non-Hispanics. Other risk factors for child overweight at 4 years of age, however—including being Hispanic, being male, having a high birthweight, having a mother with a high prepregnancy BMI, and being fed infant formula along with breastfeeding—appeared to lessen or offset the protective influence of breastfeeding for 9 weeks or more.
Although numerous studies support the protective association of increased breastfeeding duration against childhood and adolescent overweight, other studies do not. Burke et al. did not find the duration of breastfeeding to be associated with a child’s risk of being overweight between 1 and 8 years of age.17 Hediger et al. found a reduced risk of being at risk of overweight for ever-breastfed children but no reduced risk of being overweight.18 When the 2 weight categories were combined in another study, the investigators found a significant association.9 A 2006 study that used dual-energy X-ray absorptiometry to examine 5-year-old children’s body composition found neither breastfeeding nor the timing of the introduction of complementary foods to be associated with adiposity.19
We examined the relationship between breastfeeding duration and overweight in a Kansas WIC population. Other recent studies have explored a relationship between WIC and overweight in children, with particular focus on the program’s practice of providing infant formula as a mediating factor.20 The relationship of the WIC program itself, and the effect of providing formula, to later overweight status offer intriguing future research opportunities.
Our study had several limitations. Grummer-Strawn and Mei raised a question with their study that is true of ours as well.4 They noted that "women who breastfeed for longer durations could be different from women who do not breastfeed in ways not measured by the variables that we considered."4(pe85) Although our sample included all usable linked data from each mother–child pair with at least 1 WIC visit annually during the study period, selection bias could have occurred when potential study participants did not continue their WIC participation. Also, breastfeeding data available from the PedNSS, although valuable, were not detailed enough to differentiate exclusive from partial breastfeeding.
No single factor has been shown to protect a child from overweight, and accordingly, no single trait or practice can be identified as the cause of childhood overweight. Our study design provided consistency with other studies,4,9 which allows comparison and possible replication. Earlier researchers identified possible biological bases for breastfeeding’s protective effect against overweight.21
In theory, breastfeeding duration is likely 1 factor in a series of ecological influences that affect a child’s propensity for overweight. Our results indicated that children of mothers with prepregnancy BMIs in the overweight or obese categories were more likely to be overweight at age 4 years. Li et al. determined that the combined interaction between maternal prepregnancy obesity and lack of breastfeeding resulted in the greatest risk of children becoming overweight.22 Hediger et al. found that the proportion of overweight children nearly tripled with maternal overweight status and more than quadrupled with maternal obesity status.18 Our results and those of others4 showing protective effects of longer breastfeeding against overweight in non-Hispanic participants suggest that cultural and environmental factors may override those protective benefits. Additional research should examine not only breastfeeding’s effect but also the effects of ethnicity and culture, maternal prepregnancy BMI, and child’s birthweight and gender as factors influencing overweight in children.
Although it is not possible to claim a simple causal relationship between extended breastfeeding and lower risk of childhood overweight, our study adds support for an associated protective link between the two. Our results corroborate previous research that identified the protective effect of breast-feeding duration against child overweight in non-Hispanic children at 4 years of age and support the current public health emphasis in this country on increasing breastfeeding initiation and duration.
|
Acknowledgments |
|---|
The authors acknowledge the valuable contributions of Patricia Dunavan and Mary Washburn, for information about the Kansas Special Supplemental Nutrition Program for Women, Infants, and Children and assistance with accessing data, and Carol MacGowan and Ellen W. Borland, for access to participant data from the Pediatric Nutrition Surveillance and Pregnancy Nutrition Surveillance Systems via the Centers for Disease Control and Prevention. Paula Peters, Kansas State Research and Extension and Department of Human Nutrition; Valentina Remig, Department of Human Nutrition; and Ann Murray, School of Family Studies and Human Sciences, Kansas State University, Manhattan, reviewed the article and contributed valuable suggestions.
Human Participant Protection
The institutional review board of the Kansas Department of Health and Environment Bureau of Children, Youth, and Families and the committee on research involving human subjects at Kansas State University reviewed and approved the study protocol.
|
Footnotes |
|---|
Peer Reviewed
Contributors
S.B. Procter reviewed the literature, designed the study, assisted in data extraction and analysis from the files of the Centers for Disease Control and Prevention, and wrote the article. C.A. Holcomb assisted with project design, directed and supervised data extraction and analysis, provided comments for the article, and reviewed the article.
Accepted for publication March 20, 2007.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
1. Ogden CL, Flegal KM, Carroll MD, Johnson CL. Prevalence and trends in overweight among US children and adolescents, 1999–2000. JAMA. 2002; 288:1728–1732.
2. Strauss RS, Pollack HA. Epidemic increase in childhood overweight, 1986–1998. JAMA. 2001; 286:2845–2848.
3. Dietz WH, Gortmaker SL. Preventing obesity in children and adolescents. Annu Rev Public Health. 2001;22:337–353.[CrossRef][Web of Science][Medline]
4. Grummer-Strawn LM, Mei Z. Does breastfeeding protect against pediatric overweight? Analysis of longitudinal data from the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance System. Pediatrics. 2004;113:e81–e86.
5. Harder T, Bergmann R, Kallischnigg G, Plagemann A. Duration of breastfeeding and risk of overweight: a meta-analysis. Am J Epidemiol. 2005;162:397–403.
6. Arenz S, Ruckerl R, Koletzko B, von Kries R. Breast-feeding and childhood obesity—a systematic review. Int J Obes Relat Metab Disord. 2004;28:1247–1256.[CrossRef][Web of Science][Medline]
7. Armstrong J, Reilly JJ, Child Health Information Team. Breastfeeding and lowering the risk of childhood obesity. Lancet. 2002;359:2003–2004.[CrossRef][Web of Science][Medline]
8. von Kries R, Koletzko B, Sauerwald T, et al. Breast feeding and obesity: cross-sectional study. BMJ. 1999;319:147–150.
9. Heinig MJ, Dewey K. Overweight and Obesity in Children and Their Mothers: Does Breastfeeding Make a Difference? Davis, Calif: Regents of the University of California; 2004. Independent Study Module No. 020406.
10. Centers for Disease Control and Prevention, Division of Adolescent and School Health. HHS efforts to combat the obesity epidemic among children and adolescents. Available at http://www.cdc.gov/washington/testimony/Ob6162004204.htm. Accessed September 3, 2004.
11. Ryan AS, Zhou W. Lower breastfeeding rates persist among the Special Supplemental Nutrition Program for Women, Infants, and Children participants, 1978–2003. Pediatrics. 2006;117:1136–1146.
12. Abbott Laboratories, Ross Products Division. Breastfeeding trends—2002. Mothers survey. Available at: http://abbottnutrition.com/resources/en-US/home/breastfeeding/BF_Trends_2003.pdf. Accessed June 2004.
13. Centers for Disease Control and Prevention. Pediatric and Pregnancy Nutrition Surveillance System. What is PedNSS/PNSS? Available at: http://www.cdc.gov/pednss/what_is. Accessed April 20, 2006.
14. Mei Z, Scanlon KS, Grummer-Strawn LM, Freedman DS, Yip R, Trowbridge FL. Increasing prevalence of overweight among US low-income preschool children: the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance, 1983 to 1995. Pediatrics [online version]. 1998;101. Available at: http://www.pediatrics.org/cgi/content/full/101/1/e12. Accessed April 22, 2004.
15. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11. 2002 May (246):1–190.
16. Krider C, Clifford N. Demographic Changes in Kansas: Implications for State Policy. Lawrence: University of Kansas, Kansas Center for Community Economic Development, Policy Research Institute; 2001: 27–33. Technical Report Series, No. 53, prepared for the Kansas Futures Committee, Kansas Legislature.
17. Burke V, Beilin LJ, Simmer K, et al. Breastfeeding and overweight: longitudinal analysis in an Australian birth cohort. J Pediatr. 2005;147:56–61.[CrossRef][Web of Science][Medline]
18. Hediger ML, Overpeck MD, Kuczmarski RJ, Ruan WJ. Association between infant breastfeeding and overweight in young children. JAMA. 2001; 285:2453–2460.
19. Burdette HL, Whitaker RC, Hall WC, Daniels SR. Breastfeeding, introduction of complementary foods, and adiposity at 5 y of age. Am J Clin Nutr. 2006;83: 550–558.
20. Rose D, Bodor JN, Chilton M. Has the WIC incentive to formula-feed led to an increase in overweight children? J Nutr. 2006;136:1086–1090.
21. Dewey KG. Is breastfeeding protective against child obesity? J Hum Lact. 2003;19:9–18.
22. Li C, Kaur H, Choi WS, Huang TT, Lee RE, Ahluwalia JS. Additive interactions of maternal prepregnancy BMI and breast-feeding on childhood overweight. Obes Res. 2005;13:362–371.[Web of Science][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |