Objective. To determine whether a rapid rate of weight gain in early infancy is associated with overweight status in childhood.
Design. Prospective, cohort study from birth to age 7 years.
Setting. Twelve sites across the United States.
Participants. Twenty-seven thousand, eight hundred ninety-nine (27 899) eligible participants born at full term between 1959 and 1965.
Main Outcome Measure. Overweight status at age 7 years, defined by a body mass index above the 95th percentile of the Centers for Disease Control and Prevention reference data.
Results. In the 19 397 participants with complete data (69.6%), the prevalence of overweight status at age 7 years was 5.4%. The rate of weight gain during the first 4 months of life (as 100 g/month) was associated with being overweight at age 7 years, after adjustment for several confounding factors: odds ratio: 1.38; 95% confidence interval: 1.32–1.44. This association was present in each birth weight quintile, and remained significant after adjustment for the weight attained at age 1 year (odds ratio: 1.17; 95% confidence interval: 1.11–1.24).
Conclusions. A pattern of rapid weight gain during the first 4 months of life was associated with an increased risk of overweight status at age 7 years, independent of birth weight and weight attained at age 1 year. These findings may lead to new hypotheses regarding the cause of childhood obesity, which may contribute to our understanding of this increasing public health problem in the United States.
The prevalence of overweight and obesity is increasing in US adults1 and children.2 For example, between 1976–1980 and 1999, the prevalence of adult obesity increased from 14% to 26%.3,4 The prevalence of overweight status in children aged 6 to 11 years tripled from 1971–1974 (approximately 4%) to 1999 (approximately 13%). Thus, in addition to efforts to improve obesity treatment,5,6 obesity prevention has become a major public health priority in the United States.7,8
Prevention of obesity beginning in childhood is critical, as childhood obesity is associated with the development of unfavorable health outcomes during childhood.9–11 In addition, childhood obesity is not only associated with an increased risk for adult obesity, but also with unfavorable health outcomes in adulthood, independent of weight status in adulthood.9,12,13 Excessive weight in childhood often persists into adulthood,14 especially in children with overweight parents15 and is, therefore, also a risk factor for adult obesity and its complications.9
How and when to intervene to prevent obesity in childhood is a challenging question. Studies have identified several critical periods for the development of childhood- or adolescent-onset obesity.16 These include gestation,17 the period of “adiposity rebound” (ages 3–6 years),18 and puberty.16 Epidemiologic studies have identified several interrelated factors that may predispose children to obesity development during early life: maternal overweight status,15,19 maternal gestational diabetes,20 high birth weight,14,19,21,22 formula feeding rather than breastfeeding,23,24 early introduction of solid foods,25 overfeeding,26,27 and television viewing.28,29 Many of these factors have changed in the past decades, including recommendations for infant feeding.
Infancy has not been widely targeted for obesity prevention. However, some evidence suggests that the rate of weight gain during the first few months of life, and related determinants, such as type of infant feeding, may influence weight status later in childhood,23,30,31 as well as the later development of adult cardiovascular disease.32 If all of these findings were true, the first months of life should be a special focus of intervention to facilitate appropriate patterns of infant feeding and growth. The aim of this study was to assess the association between the rate of weight gain between birth and age 4 months and the prevalence of overweight status in childhood, in a cohort of US children who were followed prospectively from birth.
National Collaborative Perinatal Project (CPP)
The CPP was a multicenter, cohort study initiated in 1959 to investigate risk factors for cerebral palsy at 12 US sites: Boston Hospital for Women and Children, Hospital Medical Center, Boston, Massachusetts; Children’s Hospital, State University of New York at Buffalo, New York; Charity Hospital, New Orleans, Louisiana; Columbia-Presbyterian Hospital, New York, New York; Johns Hopkins Hospital, Baltimore, Maryland; Medical College of Virginia, Richmond, Virginia; University of Minnesota Hospitals, Minneapolis, Minnesota; Metropolitan Hospital, New York, New York; University of Oregon Medical Center, Portland, Oregon; Pennsylvania Hospital, Philadelphia, and Children’s Hospital of Philadelphia, Pennsylvania; Child Study Center, Brown University, Providence, Rhode Island; and Gailor Hospital, Memphis, Tennessee. Between 1959 and 1965, 58 760 women with singleton pregnancies presenting for prenatal care in these centers were enrolled in the study after informed consent was obtained. Among them, 6454 were lost to follow-up before delivery or experienced fetal death, resulting in a cohort of 52 306 registered live births. Maternal data were collected at enrollment using a questionnaire, which was slightly modified in 1962; hence, not all variables were available for every woman. Follow-up data were collected on the children at ages 4 months, 1 year, and 7 years.
The present study assessed the prevalence of overweight status at age 7 years as a function of the rate of weight gain during the first 4 months of life, using a prospective, cohort study design. To eliminate the potential for confounding by differences in patterns of early weight gain in infants who were preterm or postterm, the sample was restricted to the 27 889 participants with birth documented to be at term gestation. Gestational age was <37 weeks for 2507 participants, >42 weeks for 377 participants, and was unavailable for 21 533 births, mostly from participants born before 1962, when this information started to be widely recorded as part of the research protocol. Maternal and familial variables of interest were those with theoretical or previously reported influence on childhood weight status. Maternal prepregnancy weight, maternal education, race, number of siblings in the household, and number of previous pregnancies were recorded by interview at enrollment of the pregnant woman into the CPP. Maternal height was measured at time of delivery and used to calculate prepregnancy body mass index (BMI) as reported prepregnancy weight in kg divided by height in meters squared.
Variables of interest for the children included sex, gestational age, birth order, weight, and length at birth and at the subsequent follow-up visits. According to the CPP study protocol, gestational age was determined by “the senior physician assuming responsibility for the delivery using all clinical and historical information.” The participants were considered as first-born if the mother reported “no previous pregnancy,” or “no children living in the household” at the time of enrollment, depending on which question was asked at that time. The method of infant feeding was also of interest, but only initial method of infant feeding, in the first few days after birth, was documented for a subsample of 9194 participants. Furthermore, this information was widely recorded for the study participants at some study sites (up to 99%), but less so at other sites (as low as 12%). Therefore, this variable was only used in a secondary analysis for this report.
The main exposure variable, monthly rate of weight gain during the first 4 months of life, was calculated as weight at the 4-month examination minus birth weight divided by exact age in months at the 4-month examination and was expressed as 100 g per month. The main outcome, overweight status at the 7-year examination, was defined using the presently recommended definition for the US population,33 corresponding to an age-specific (1-month precision) and sex-specific BMI above the 95th percentile of the Centers for Disease Control and Prevention growth charts. The analysis was repeated using the alternative definition for obesity from the International Obesity Task Force34 based on age-specific (half-year precision) and sex-specific BMI cutoff points corresponding to an adult BMI ≥30 kg/m2. The actual age of the child at the 7-year examination was used for assessing these 2 outcomes, rather than age 7 for every participant.
Most variables were skewed and are, therefore, described using proportions or medians, 2.5th and 97.5th percentiles. Participants with complete and incomplete follow-up were compared using χ2 or Wilcoxon rank sum test, as appropriate. The unadjusted associations between the main exposure or the confounding variables and childhood overweight status were estimated by simple logistic regression or χ2, as appropriate, and expressed as the odds ratio (OR) and 95% confidence intervals (CI). As the outcome (overweight status) was relatively infrequent, OR gave a relatively good estimate of relative risks. The population attributable risk percent was calculated as the difference in prevalence between the exposed and the unexposed groups, divided by the prevalence in the exposed group, and multiplied by 100. Possible interactions in the association of the main exposure with the main outcome were explored using stratified analyses, including Mantel-Haenszel test of homogeneity, for the following variables with conceptual or previously reported influence: sex, race, birth weight, birth order, and method of infant feeding. To adjust for additional explanatory variables while exploring the association between the main exposure and the main outcome, a multiple logistic regression model was constructed that included the potential confounding variables identified in the unadjusted analysis (P < .1), plus gestational age, exact age at assessment, and location of the study site. An additional model that included the weight attained at age 1 year adjusted the association between the rate of weight gain during the first 4 months of life and overweight status at age 7 years for the weight attained in the interval between these 2 measurements. Furthermore, another alternative model was also constructed using weight gain between 4 and 12 months instead of weight attained at age 1 year. To assess the potential influence of breastfeeding on this relationship, the analysis was repeated in the subset of participants with documented initial infant feeding method. In all these analyses, the main exposure variable, weight gain during the first 4 months of life, was expressed in 100 g per month. Therefore, an OR of 1.1 would approximately correspond to a 10% increased risk for overweight status at age 7 years for each 100 g per month increase in weight gain between birth and age 4 months. All statistical tests were 2-tailed and a value of P < .05 considered statistically significant. The analysis was conducted using Stata 6.0 (Stata Corp, College Station, TX).35
Among the 27 889 eligible participants born at term gestation, 19 397 (69.6%) were measured at birth, 4 months, and 7 years (Table 1). The participants with complete and incomplete measurements were similar in sex distribution and gestational age at birth. However, children with complete measurements were more likely than those with incomplete measurements to be black (50.7% vs 47.4%, respectively; P < .001), to have a higher birth weight (3240 g vs 3210 g; P < .001), an older sibling (75.4% vs 70.1%; P < .001), and a mother with a higher BMI (22.9 kg/m2 vs 22.6 kg/m2; P < .001) and with a higher education (10.9 years vs 10.5 years; P < .001).
The unadjusted association between the rate of weight gain during the first 4 months of life (per 100 g per month) and overweight status at age 7 years was OR: 1.29; 95% CI: 1.25–1.33 (Table 2). This means that for each 100 g per month increase in weight gain between birth and age 4 months, the risk for overweight status at age 7 years was increased by 29% (95% CI: 25%–33%). The population attributable risk percentage of overweight status for the highest weight gain quintile was 21.0%. The other variables associated with an increased risk of overweight status in childhood in the unadjusted analysis were: female sex, white race, first-born status, birth weight, weight at age 1 year, maternal BMI, and a lower maternal education (Table 2).
As the effect of weight gain in the first 4 months of life on the risk of childhood obesity may be different at various birth weights, the participants were stratified by birth weight quintiles, and no significant interaction was observed (test for heterogeneity; P = .5; Fig 1). In similar analyses, there was no significant interaction in the association between early infancy weight gain and childhood overweight status with sex, race, and first-born status (test for heterogeneity; P > .1 in all cases).
A multiple logistic regression model was used to adjust for potential confounding variables of the association between early weight gain and childhood overweight status (Table 3, Model 1). This analysis confirmed an independent association between the rate of weight gain in the first 4 months of life and overweight status 7 years later (OR: 1.38; 95% CI: 1.32–1.44), corresponding to a 38% (95% CI: 32%–44%) increased risk for each 100 g per month increase in early infancy weight gain. To test the possibility that this association was simply explained by persistence of excessive weight, we included weight at age 1 year in an alternative analysis (Table 3, Model 2). Although the magnitude of the association between the rate of weight gain during the first 4 months of life and overweight status in childhood decreased, it remained independently significant, regardless of the weight achieved at age 1 year (OR: 1.17; 95% CI: 1.11–1.24). An additional logistic regression model, using weight gain between 4 and 12 months, rather than weight attained at age 1 year, gave similar results: Weight gain from birth to 4 months old (OR: 1.34; 95% CI: 1.28–1.41), weight gain between 4 and 12 months (OR: 1.60; 95% CI: 1.48–1.73). It should be emphasized that these 2 ORs cannot be directly compared. The duration of the 2 periods and the physiologic rate of growth during these periods are different; therefore, a sustained increase of 100 g per month has a different meaning for the 2 periods.
All components of this analysis were repeated using the alternative definition of obesity of the International Obesity Task Force,34 and the results were essentially the same (not shown). The analysis was repeated in the subset of 5479 participants with complete data and some information on initial infant feeding method. Initiation of breastfeeding was not significantly associated with overweight status at age 7 years when compared with no breastfeeding (OR: 1.28; 95% CI: 0.87–1.88), and no interaction was detected for initiation of breastfeeding in the association of early rate of weight gain and childhood overweight status. When initiation of breastfeeding was introduced into the final model (Model 2), the results for the other variables were essentially the same (not shown).
The present study showed that a more rapid rate of weight gain during the first 4 months of life was associated with an increased risk for childhood overweight status in a large, diverse cohort of US children. This association seemed to be independent of birth weight and of the weight attained at age 1 year.
The association of early rate of weight gain and childhood overweight status was clinically and statistically significant: for each 100 g weight gain increase per month, the risk for overweight status 7 years later was increased by approximately 30%. The participants with the most rapid weight gain during the first 4 months of life were more likely to become overweight children than those with the slowest rate of weight gain. Furthermore, nearly 20% of the overweight status at age 7 years could be attributed to having a rate of weight gain during the first 4 months of life in the highest quintile. In this cohort of children born at full-term, and mostly with a weight appropriate for gestational age, the association of early infancy weight gain and childhood overweight status was independent of birth weight, and was of similar magnitude across birth weight quintiles. It could be argued that our findings may be explained by the phenomenon of tracking of excessive weight over time. However, if this were the case, statistical adjustment for the weight achieved at age 1 year, which was on the pathway of excessive weight persistence, would have resulted in the disappearance of this association. Although weight at age 1 year was strongly associated with the outcome and the magnitude of the association between early infancy rate of weight gain and childhood overweight status decreased after adjustment, this association remained clinically and statistically significant. Each 100 g of weight gain increase per month in the first 4 months of life was associated with a modest increase of 17% in the risk for overweight status 7 years later, independent of the weight achieved at age 1 year.
We therefore propose that early infancy also constitutes a critical period for the establishment of obesity, in addition to fetal life, the period of “adiposity rebound,” and puberty. Early infancy is physiologically a period of very rapid weight gain, with birth weight doubling in the first 4 to 6 months of life.22 This represents the greatest proportional weight gain in the postnatal life cycle and, therefore, may correspond to a critical period for the development of obesity and energy balance regulation mechanisms. This hypothesis is supported by previous smaller epidemiologic studies30,31 and animal models, describing permanent changes in brain structure36 and enzymatic function37 associated with adult obesity in animals with a rapid neonatal weight gain. Alternatively, early weight gain may reflect the beginning influence of a genetic predisposition to overweight, rather than environmental factors or feeding patterns.
Low birth weight38 and a rapid weight gain in childhood32 have also been associated with an increased risk for cardiovascular disease in adulthood. Full-term infants born with a low birth weight are more likely to gain weight rapidly early in life, as part of the usual pattern of catch-up growth. A rapid rate of early weight gain may thus be a common factor between the development of obesity and cardiovascular disease later in life. These 2 conditions often cluster in individuals.
The present results also confirm the well-described association of childhood overweight status with maternal BMI, lower maternal education, and high birth weight. The increased risk for overweight status in first-born children, compared with children with older siblings, also agrees with previous results from a subsample of black participants issued from the same cohort who were followed to young adulthood.39 In this cohort of participants born in the 1960s, black children were less likely to be overweight than were white children, whereas currently black children are more likely than white children to be overweight.2 In our study, however, this racial difference disappeared after adjustment for other important confounding variables. In the small subsample of participants with some information on infant feeding mode, initiation of breastfeeding was not significantly associated with a decreased risk for excessive weight, as has been observed in other studies.23–25 Furthermore, the effect of early weight gain on overweight status at age 7 years seemed to be similar in participants with some or with no breastfeeding. These results must, however, be interpreted with caution, as, in the present study, information on initiation of breastfeeding was widely available only at some of the sites, the duration of breastfeeding was not assessed, and the broad confidence interval did not exclude a possible existing association. Furthermore, initiation of breastfeeding does not always reflect duration or exclusive breastfeeding, which may be important in the association between early weight gain and overweight status later in life.
This study had some limitations. First, not all the children initially enrolled in the study completed all measurements; 69.6% were completed. However, this loss to follow-up would only create a biased estimate of the association between early weight gain and the risk for childhood overweight status if the association of interest were different between the participants with complete and the participants with incomplete follow-up. Furthermore, the differences between these 2 groups were of small magnitude (eg, a 30-g difference in birth weight) although, because of the large sample size, they were statistically significant. The impact of these small differences is difficult to assess, but the differences are unlikely to bias the observed associations enough to change the conclusions of the study. It is unclear whether the association between early rate of weight gain and overweight status at age 7 years, described in these participants born in US cities in the 1960s, is still valid today, and if it would be present in other environments, such as rural areas or countries other than the United States. Recommendations for infant feeding have changed since the 1960s, and introduction of solid foods at an early age, which was more common at that time, may explain some of the early weight gain in the present study. Therefore, the study of contemporary cohorts in various environments is warranted. It is possible that parents who overfeed their infants in early infancy may also be more likely to overfeed them later in childhood, thus explaining the observed association. The design of this study could not exclude this potential confounding factor. Furthermore, the possibility that the weight gain during the first 4 months of life may reflect the beginning influence of genetic factors rather than feeding pattern could not be assessed in the present study. Weight gain during the first 4 months of life may not be linear, and a rapid rate of weight gain during the first few weeks of life may be compensated by a slower rate in the following months. The impact of the patterns of weight gain within the first 4 months of life could not be explored in the present study. Finally, it should be emphasized that the 7-year follow-up of this study is an insufficiently long duration from which to draw meaningful conclusions on the long-term impact of early weight gain. Overweight at age 7 years is not as strong a predictor of overweight in adulthood as overweight in older children or adolescents. At this time, there are no effective and safe intervention strategies in infancy for the prevention of later development of obesity.
Our study also had unique strengths. All the data were collected prospectively, as part of a research protocol investigating other health outcomes, thus limiting the potential for selection and reporting bias. Although this sample was not fully representative of the entire US population, it was conducted in 12 study sites across the country and included both black and white children.
This study demonstrated an increased risk of being overweight in 7-year-old children with a rapid rate of weight gain during the first 4 months of life. If this association is confirmed by other studies, the present results may lead to new approaches for the prevention of childhood obesity, aiming at an optimization of healthy nutrition and growth in infancy. The American Academy of Pediatrics emphasizes that exclusive breastfeeding is ideal nutrition and sufficient to support optimal growth and development in the first 6 months of life.40
This study was supported by the Nutrition Center of The Children’s Hospital of Philadelphia.
We thank Solomon H. Katz, PhD, and Jim Coleman for their support on accessing and managing the CPP data, and Justine Shults, PhD, for statistical support.
- Received May 7, 2001.
- Accepted August 22, 2001.
- Address correspondence to Nicolas Stettler, MD, MSCE, Division of Gastroenterology and Nutrition, Room 7410, The Children’s Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104-4399. E-mail:
- ↵Prevalence of Overweight and Obesity Among Adults. United States, 1999. Available at: http://www.cdc.gov/nchs/products/pubs/pubd/hestats/obese/obse99tab2.htm Accessed March 30,2001
- ↵National Heart, Lung, and Blood Institute clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. NIH Consens Statement.1998;98– 4083
- ↵Hill JO, Trowbridge FL. Childhood obesity: future directions and research priorities. Pediatrics.1998;101 :570– 574
- ↵Healthy People 2010. Nutrition and Overweight. Available at: http://www.health.gov/healthypeople/document/html/volume2/19nutrition.htm Accessed March 30,2001
- ↵National Action Plan on Overweight and Obesity [press release]. Available at: http://www.surgeongeneral.gov/todo/pressreleases/obesitypressrelease.htm Accessed March 30,2001
- ↵Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord.1999;23(suppl 2) :S2– S11
- Rosenbloom AL, Joe JR, Young RS, Winter WE. Emerging epidemic of type 2 diabetes in youth. Diabetes Care.1999;22 :345– 354
- ↵Must A. Morbidity and mortality associated with elevated body weight in children and adolescents. Am J Clin Nutr.1996;63 :445S– 447S
- ↵Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics.1999;103 :1175– 1182
- ↵Dietz WH. Critical periods in childhood for the development of obesity. Am J Clin Nutr.1994;59 :955– 959
- ↵Rolland-Cachera MF, Deheeger M, Bellisle F, Sempe M, Guilloud-Bataille M, Patois E. Adiposity rebound in children: a simple indicator for predicting obesity. Am J Clin Nutr.1984;39 :129– 135
- ↵Curhan GC, Willett WC, Rimm EB, Spiegelman D, Ascherio AL, Stampfer MJ. Birth weight and adult hypertension, diabetes mellitus, and obesity in US men. Circulation.1996;94 :3246– 3250
- ↵Pettitt DJ, Nelson RG, Saad MF, Bennett PH, Knowler WC. Diabetes and obesity in the offspring of Pima Indian women with diabetes during pregnancy. Diabetes Care.1993;16(suppl 1) :310– 314
- ↵Curhan GC, Chertow GM, Willett WC, et al. Birth weight and adult hypertension and obesity in women. Circulation.1994;94 :1310– 1315
- ↵Seidman DS, Laor A, Gale R, Stevenson DK, Danon YL. A longitudinal study of birth weight and being overweight in late adolescence. Am J Clin Nutr.1991;145 :782– 785
- ↵von Kries R, Koletko B, Sauerwald T, et al. Breast feeding and obesity: cross sectional study. BMJ.1999;319 :147– 150
- ↵Neumann CG, Alpaugh M. Birthweight doubling time: a fresh look. Pediatrics.1976;57 :469– 473
- ↵Stunkard AJ, Berkowitz RI, Stallings VA, Schoeller DA. Energy intake, not energy output, is a determinant of body size in infants. Am J Clin Nutr.1999;69 :524– 530
- ↵Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. Relationship of physical activity and television watching with body weight and level of fatness among children: results from the Third National Health and Nutrition Examination Survey. JAMA.1998;279 :938– 942
- ↵Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger DB. Association between postnatal catch-up growth and obesity in childhood: prospective cohort study. BMJ.2000;320 :967– 971
- ↵Eid EE. Follow-up study of physical growth of children who had excessive weight gain in first six months of life. BMJ.1970;2 :74– 76
- ↵Eriksson JG, Forsén T, Tuomilehto J, Winter PD, Osmond C, Barker DJP. Catch-up growth in childhood and death from coronary heart disease: longitudinal study. BMJ.1999;318 :427– 431
- ↵Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC Growth Charts: United States. Advance Data, Vital and Health Statistics. No. 314. Hyattsville, MD: National Center for Health Statistics; 2000. Available at: http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/fullreport.htm Accessed August 20,2001
- ↵Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ.2000;320 :1240– 1243
- ↵StataCorp. Stata Statistical Software, Release 6.0. College Station, TX: Stata Corporation;1999
- ↵Stettler N, Tershakovec AM, Zemel BS, et al. Early risk factors for increased adiposity: a cohort study of African American subjects followed from birth to young adulthood. Am J Clin Nutr.2000;72 :378– 383
- ↵American Academy of Pediatrics, Work Group on Breastfeeding. Breastfeeding and the use of human milk. Pediatrics.1997;100 :1035– 1039
- Copyright © 2002 by the American Academy of Pediatrics