Association Between High Birth Weight and Later Central Obesity in 9-Year-Old Schoolchildren

Abstract

Background and Objective:

Studies have suggested that birth weight (BW) is associated with body mass index (BMI), but its association with waist circumference (WC) in children should be further explored. To determine the association between central obesity (OB) in 9-year-old Argentinean schoolchildren and high BW.

Methods:

Schoolchildren (n = 2567, 1157 males) aged 8.7 ± 2.1 years from 10 elementary schools in 5 states in Argentina were examined between April 2017 and September 2019. Mothers submitted children's BW information. Pediatricians assessed anthropometric measures and blood pressure (BP). Central OB was defined for children as WC ≥90th percentile for age and gender.

Results:

The prevalence of overweight (OW) and OB (OW/OB) was 42.7% (1095) and that of central OB was 34.8% (856) in 9-year-old children. The prevalence of low BW (<2500 grams) and high BW (>4000 grams) was 6.6% (n = 169) and 7.4% (n = 190), respectively. BW (3.25 vs. 3.36 kg), weight (31.38 vs. 42.88 kg), BMI (17.29 vs. 22.25 kg/m²), BMI z-scores (z-BMI; 0.25 vs. 1.63), systolic BP (96 vs. 98 mmHg), and diastolic BP (59 vs. 60 mmHg) were significantly lower in 9-year-old children without central OB than in those with central OB, respectively. Multiple logistic regression analysis using central OB as the dependent variable showed that high BW [odds ratio, 1.98 (95% confidence interval 1.44–2.73)] was associated with central OB, adjusted for age, gender, and systolic and diastolic BP.

Conclusion:

This study shows that central OB in 9-year-old children was associated with high BW. Future longitudinal studies should be performed to confirm this finding. Clinical Registration number, IATIMET-08102019.

Introduction

Obesity (OB) and the cardiometabolic complications associated with OB are observed more and more frequently among children.^1,2 The prevalence of OB increased from 6.5% to 18.4% in schoolchildren aged 6–11 years from 1980 to 2017 in the United States.¹ A multicenter study from four areas of Argentina performed by our group in more than 1200 schoolchildren in 2019 showed that the prevalence of overweight (OW) and OB (OW/OB) was 42.4%, which is very similar to that of the United States.^1,2 Obese children are more likely to become obese adults and to have a higher risk of cardiometabolic diseases.³ The origins of the OB epidemic are linked with the first phase of children's lives. Regarding the association of OB and birth weight (BW), the existing data seem to be conflicting. The thrifty phenotype hypotheses have proposed that low BW was a risk factor for the progress of OW/OB and cardiometabolic diseases, probably due to the effects of poor nutrition and subsequent impairment in glucose–insulin metabolism.⁴ Furthermore, a British study found that low BW was associated with fat centralized deposits in adolescent girls.⁵

However, the focus of most research was on low BW without considering individuals with high BW.^4,5 In contrast, our group's previous study showed that high, rather than low, BW was related to OW/OB and cardiometabolic markers.⁶ Children with high BW are commonly exposed to maternal OB with an adverse intrauterine environment, increasing cardiometabolic risk.⁷ Consistently, a meta-analysis showed that high BW was associated with a greater than twofold the risk of developing OW/OB, irrespective of ethnicity, gender, socioeconomic class, and maternal weight.⁸ However, body mass index (BMI) is a surrogate measure for total adiposity and does not describe body fat distribution, which may be strongly associated with cardiometabolic complications than total adiposity.⁹

As far as we know, few studies have widely assessed the associations between central fat distribution in Argentinean schoolchildren and neonatal adiposity. The objective of this study was to determine the association of central OB and high BW in 9-year-old Argentinean schoolchildren.

Methods

This cross-sectional study was performed in four areas distributed throughout Argentina between April 2017 and September 2019. The human rights committee of the University of Buenos Aires approved the study. Every child and guardian submitted informed consent in writing after the details of the study were described.

This study is part of a multicenter study, Cariño (cardiologia en niños), performed in different areas of Argentina. The study included four locations distributed within Argentina: central, western central, northwestern, and southern Argentina. The main purpose of the Cariño project was to explore the association between BW and age, gender, BMI, waist circumference (WC), and blood pressure (BP) in 9-year-old children.

For the selection of the sample of children, a two-stage cluster sampling was applied. The sample framework comprised the schools listed in each area, and consideration was taken of the number of students in each school. In the first stage, the schools were selected using the Sampford proportional to size sampling method, which was programmed in R. The number of schools selected was previously determined so that the number of students included would ensure an error of ∼5% in each area, considering the correction factor for the finite population. Children were selected randomly within schools. All elementary grades were included (age range 5–13 years old). Given the fact that in a previous study, the prevalence of OW/OB was ∼42% among children,² the sample size was estimated to achieve an error <0.05.

We examined 2567 children, which means that the global sample error was <0.02. The sample was sufficiently robust to represent the regions included. The subjects were categorized into socioeconomic levels based on factors including education completed and whether or not the home had a refrigerator or dirt floor. These two indicators are used to identify families of low socioeconomic status by the National Statistics and Censuses Institute of Argentina.¹⁰ Mothers submitted children's BW information. A pilot group of 50 mothers was asked to bring the child's health tracking book where the BW information is recorded. There was a high concordance level between the maternal report and registered information of the BW (98.3%), which is consistent with the validity of the self-report for BW.¹¹ BW was divided into three categories: low (<2500 grams), normal, and high BW (>4000 grams).

Pediatricians assessed anthropometric measures and BP. Anthropometric measures were performed with children wearing light clothing and no shoes. BMI was determined using weight (in kilograms) divided by height (in meters) squared. BMI z-scores (z-BMI) were also determined.¹² Students were grouped as underweight (<5th percentile), normal weight (5th to <85th percentile), OW (85th to <95th percentile), or obese (≥95th percentile) following the norms set by the U.S. Centers for Disease Control and Prevention (CDC).¹² The WC measurement was performed at the level of the umbilicus and was recorded at 0.1 cm. A flexible nonelastic tape measure was used with the child standing without clothing covering the waist area. Central OB was defined for children as WC ≥90th percentile for age and gender based on 3,000 normal Argentinean children.¹³

BP was measured during physical examination in a standardized manner.¹⁴ The overall individual response rate from the whole group was 91%. Exclusion criteria were as follows: (1) missing anthropometric measurements, (2) known diabetes or other chronic disease, and (3) the use of medication that would affect BP. Subjects selected in our sample were found to have no significant difference in socioeconomic status, gender, or age, with those excluded because of missing data.

Statistical analysis

Categorical variables were disclosed as numbers (n) and percentages (%), and were compared using the chi-squared test. Means and standard deviations were presented for the normally distributed continuous variables. Continuous variable distribution normality was tested by applying the Kolmogorov–Smirnov test. A t test was utilized to analyze mean values of normally distributed variables across the groups. Variables with a skewed distribution were logarithmically transformed for observation. Bonferroni's adjustment was utilized when various comparisons were carried out. For the analysis of variance test, homogeneity of variances was tested. When it was not validated, Brown Forsythe test was performed. Multiple logistic regression analyses were used to find the association between central OB and gender, age, BP, and BW in 9-year-old children. The software package SPSS version 22 was used to perform statistical analyses. P values <0.05 were considered to be statistically significant.

Results

Schoolchildren (n = 2567, 1157 males) aged 8.8 ± 2.1 years were examined (Table 1). Socioeconomic level: none of the families had a dirt floor, 3.2% did not have a refrigerator, and 73.6% of the parents had a high school education or higher.

Table 1.

Clinical Characteristics According to Gender

	Males (N = 1157, 45.1%)	Females (N = 1410, 54.9%)	Total (N = 2567)
Age (years)	8.81 ± 2.03	8.69 ± 2.15	8.74 ± 2.09
BW (kg)	3.37 ± 0.55	3.22 ± 0.58^a	3.29 ± 0.57
Weight (kg)	36.13 ± 13.2	34.83 ± 12.57^a	35.41 ± 12.88
Height (m)	1.35 ± 0.13	1.34 ± 0.14	1.34 ± 0.14
BMI	19.29 ± 4.26	18.85 ± 3.92^a	19.05 ± 4.08
Waist (cm)	66.95 ± 11.54	65.31 ± 10.58^a	66.02 ± 11.03
z-BMI	0.8 ± 1.17	0.67 ± 1.02^a	0.73 ± 1.09
Systolic BP (mmHg)	96.62 ± 18.32	96.25 ± 17.91	96.46 ± 18.07
Diastolic BP (mmHg)	59.35 ± 13.93	59.72 ± 13.62	59.59 ± 13.74

Significance: P < 0.05.

BMI, body mass index; BP, blood pressure; BW, birth weight; CDC, Centers for Disease Control and Prevention; SD, standard deviation; z-BMI, BMI z-score.

Clinical and metabolic characteristics, according to gender, are depicted in Table 1. Students were grouped according to gender. There was not a significant difference in age or BP between genders. Boys were born 0.15 kg heavier than girls (Table 1). Boys had significantly higher weight, BMI, and WC than girls. The prevalence of OW/OB was 42.7% (1095) and that of central OB was 34.8% (856). Children were divided into two groups according to the presence of central OB (Table 2). Children with central OB were younger than those without central OB. Weight, height, BMI, z-BMI, BW, and systolic and diastolic BP were significantly higher in children with central OB than in those without central OB (Table 2).

Table 2.

Clinical Characteristics According to the Presence of Central Obesity

	WC <90th percentile, N = 1606 (65.2%)	WC ≥90th percentile, N = 856 (34.8%)
Age (years)	8.83 ± 2.08	8.59 ± 2.11^a
BW (kg)	3.25 ± 0.56	3.36 ± 0.59^a
Weight (kg)	31.38 ± 9.46	42.88 ± 14.92^a
Height (m)	1.33 ± 0.14	1.37 ± 0.13^a
BMI	17.29 ± 2.45	22.25 ± 4.32^a
WC (cm)	60.74 ± 6.8	75.92 ± 10.68^a
z-BMI	0.25 ± 0.93	1.63 ± 0.77^a
Systolic BP (mmHg)	96.4 ± 17.53	97.61 ± 18.55^a
Diastolic BP (mmHg)	59.62 ± 13.52	60.3 ± 13.88^a

Significance: P < 0.05.

WC, waist circumference.

Children were divided into three groups according to the presence of low, normal, and high BW (Table 3). The prevalence of low (<2500 grams) and high BW (>4000 grams) was 6.6% (n = 169) and 7.4% (n = 190), respectively. There was no significant difference in age or BP in 9-year-old children between the low BW, the normal BW, and the high BW groups. Mean values of weight, height, and WC in 9-year-old children were significantly higher in the high BW versus the low BW groups, and in the high BW versus the normal BW groups. BMI and z-BMI values were significantly different between all groups, which indicates that BMI increased significantly as BW increased (Table 3).

Table 3.

Clinical Characteristics According to Birth Weight

	Low BW, N = 169	Normal BW, N = 2208	High BW, N = 190
Age (years)	8.83 ± 2.12	8.69 ± 2.12	8.89 ± 2.1
BW (kg)^a	2.04 ± 0.44	3.30 ± 0.34	4.25 ± 0.28
Weight (kg)^b,c	33.36 ± 11.75	35.21 ± 12.92	38.83 ± 13.33
Height (m)^b,c	1.34 ± 0.13	1.34 ± 0.14	1.37 ± 0.13
z-Height^b,d	−0.92 ± 8.4	0.19 ± 3.77	0.85 ± 1.33
Waist (cm)^b,c	64.45 ± 9.72	65.83 ± 11.09	69.89 ± 11.49
BMI^a	18.19 ± 3.4	18.97 ± 4.01	20.08 ± 4.04
z-BMI^a	0.44 ± 1.12	0.73 ± 1.07	1.05 ± 1.1
Systolic BP (mmHg)	96.94 ± 20.62	96.58 ± 18.03	98.12 ± 18.22
Diastolic BP (mmHg)	59.88 ± 15.19	59.52 ± 13.73	59.79 ± 12.9

Data are presented as mean ± SD. Z-score is a quantitative measure of the deviation of a specific variable taken from the mean of that population. CDC z-BMI takes into account age and gender. P values compare levels between genders. Bonferroni's adjustment was carried out as many comparisons were performed. Significance: ^aall categories are significantly different; ^blow BW versus high BW; ^cnormal BW versus high BW; ^dlow BW versus normal BW.

There was a univariate association between WC and BMI (r = 0.82; P < 0.01), BW (r = 0.11; P < 0.01), systolic BP (r = 0.19; P < 0.01), and diastolic BP (r = 0.15; P < 0.01). Multiple logistic regression analyses using central OB as the dependent variable showed that high BW [odds ratio (OR), 1.98 (95% confidence interval, CI 1.44–2.73)] was associated with central OB adjusted for age, gender, and systolic and diastolic BP. When OW/OB was used as the dependent variable, high BW [OR, 1.56 (95% CI 1.14–2.13)] was also associated with OW/OB adjusted for confounding variables (Table 4). No significant association was found between central OB or OW/OB and low BW.

Table 4.

Multiple Logistic Regression Analyses

	B	Standard error	Significance	OR	95% CI OR
Dependent variable: central OB
High BW	0.69	0.16	<0.01	1.98	1.44	2.73
Age (years)	−0.08	0.02	<0.01	0.93	0.89	0.97
Gender	0.52	0.10	<0.01	1.69	1.39	2.04
Systolic BP (mmHg)	0.01	0.01	0.06	1.01	1.00	1.02
Diastolic BP (mmHg)	0.00	0.01	0.89	1.00	0.99	1.01
Dependent variable: OW/OB
High BW	0.45	1.00	<0.01	1.56	1.14	2.14
Age (years)	−0.03	0.02	0.17	0.97	0.93	1.01
Gender	−0.29	0.09	<0.01	0.75	0.63	0.89
Systolic BP (mmHg)	0.01	0.00	<0.01	1.01	1.01	1.02
Diastolic BP (mmHg)	0.00	0.01	0.89	1.00	0.99	1.01

Bold values are significant. P < 0.001.

CI, confidence interval; OB, obesity; OR, odds ratio; OW, overweight.

Discussion

Central OB was associated with high BW, a commonly used prenatal growth marker, in 9-year-old Argentinean schoolchildren. In contrast, central OB was not significantly associated with low BW. It is of interest that this study included schoolchildren from different areas of Argentina (central, southern, western, and northwestern), which increases the influence of BW on later cardiometabolic complications in Argentinean schoolchildren. As far as we know, there are no studies of high BW and its association with central OB in a large group of Argentinean schoolchildren. Thus, our results suggested that neonatal adiposity could be a marker of future central OB. However, longitudinal studies should be performed to confirm this finding.

The “fetal origins hypothesis” supports the relationship between reduced fetal growth and later increased central adiposity.⁴ Furthermore, low BW was associated with increased central adiposity in 7- to 12-year-old children.¹⁵ Rolfe et al. showed that BW was inversely associated with visceral fat.¹⁶ However, contradictory results have been described in the association of low BW and future adiposity.^4,17 A U-shaped relationship between BW and BMI, WC, and subscapular skinfold measurements was also described.¹⁸

The inverse relationship reported between low BW and later OB may be overestimated and may reflect the impact of random error, the priority of specific results, and an inappropriate adjustment for potential confounding factors. A study performed by our group⁶ and this study did not find significant associations between OB and low BW. We found that 9-year-old children with low BW had significantly lower body fat distribution than those with high BW. The multiple regression analysis confirmed that there was no significant association between central OB and low BW.

A critical review demonstrated that high BW was associated with an increased risk of OB in childhood and adulthood.¹⁵ Furthermore, a South American study performed in Brazilian adolescents showed that high BW as compared with normal BW was associated with higher prevalence of OW/OB.¹⁹ Accordingly, our group previously found a significant association between high BW and OW/OB as well as the metabolic syndrome in children from Buenos Aires.⁶ A follow-up study of ∼1000 U.S. children showed a constant relationship between neonatal adiposity and BMI through 6 years of age in which the relationship did not change over time.¹⁷ However, a limitation of this study was the lack of measurement of body fat distribution. Body fat distribution is a main contributor to cardiometabolic risk.²⁰

The BMI has high specificity but low sensitivity to detect excess fat, and >25% of children with excess fat mass will be misclassified.²¹ A 30-year Australian report (1985–2015) found that even though OB prevalence in primary school children remained stable, the prevalence of central OB almost doubled during this period.⁹ The mechanisms related to the larger WC in subjects are unclear but might be associated with high secretory adipose tissue or increased responsiveness to the effects of adipokines.²² Furthermore, children with higher WC have higher intra-abdominal, liver, and muscle fat deposition, and, therefore, higher insulin resistance.²³

A study performed in >28,000 U.S. children aged 0–15 years in the National Health and Nutrition Examination Survey (NHANES) found that an increase in BW was significantly associated with a higher risk of central OB.²⁴ Consistently, this study showed that 9-year-old children with central OB had increased BW than those without central OB. This suggests that excessive adiposity during fetal life could lead to future cardiometabolic alterations.^19,25 In addition, genetic and epigenetic variation might be also associated with central OB.²⁶ However, limited genome-wide integrative analyses have been performed yet.²⁶ Our data can help identify children who might benefit from lifestyle interventions to prevent future cardiometabolic disease. However, prospective studies should be performed to confirm this finding.

This study involved a few limitations. It is a cross-sectional study, and we cannot determine a cause–effect relationship. Because of economic limitations, BP measurements were assessed by pediatricians only at one visit during physical examinations in each area.¹⁴ Children with high BW are a heterogeneous group that comprises children from maternal gestational diabetes, or maternal OB, or normal large children with BW, according to their genetic backgrounds. Because of lacking information, no adjustments were made for alimentary patterns, family histories of hypertension, pubertal status, or other potentially confounding factors. Another limitation is the extent to which findings are generalizable to other populations.

Despite the limitations already mentioned, the strengths of our study included a high response rate and a large number of apparently normal schoolchildren in different areas of Argentina. For adiposity measurements, we used not only BMI but also WC that accounts for the central distribution. Furthermore, we used multiple regression models to make the required adjustments for the confounding variables.

Conclusion

Central OB was observed to be significantly associated with high BW in 9-year-old children. However, low BW was not significantly associated with either central OB or OW/OB. The OB epidemic roots must be traced back to birth to develop effective OB prevention programs and avoid future cardiometabolic complications. Early interventions to reduce central OB and OW/OB in children with high BW could decrease later cardiometabolic diseases. The encouragement of healthy lifestyle behaviors for preventing and treating central OB in children and their parents, especially those with high BW, may be of importance to prevent cardiovascular disease. However, further prospective studies should be performed.

Footnotes

Acknowledgments

Members of the Cariño Study Group Collaborators are Valeria Calzia, Romina Di Firma, Ornella Lione, Gianluca Sansoni, Agustina Lizazu, Patricia Bocco, María Carolina Albornoz, Alejandra Verónica, Alejandra Burgarello, Lorena Beatriz Benitez, Dra. Silvina Radio, Carolina Beatriz Yulán, Silvana Beatriz Cáceres, Mariana Andrea De La Rosa, Martin Sabrina, Pomilio Julieta, Raina Belen, Marquez Pilar, Pareja Yamil, Debiasi Antonella, Muzaber Martin, Pipistrelli Gina, Pons Carla, and Lang Diana.

Author Disclosure Statement

There are no potential conflicts of interest.

Funding Information

No funding was received for this work.

References

Hales

, Carroll

, Fryar

, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief, 2017; 288:1–8.

Hirschler

, Lapertosa

, Scaiola

, et al. Adiposity, lifestyle behaviors, and cardiometabolic markers in Argentinean schoolchildren. Clin Chim Acta, 2020; 507:280–285.

Kumar

, Kelly

. Review of childhood: From epidemiology, etiology, and comorbidities to clinical assessment and treatment. Mayo Clin Proc, 2017; 92:251–265.

Hales

, Barker

. The thrifty phenotype hypothesis. Br Med Bull, 2001; 60:5–20.

Barker

, Robinson

, Osmond

, et al. Birth weight and body fat distribution in adolescent girls. Arch Dis Child, 1997; 77:381–383.

Hirschler

, Bugna

, Roque

, et al. Does low birth weight predict/overweight and metabolic syndrome in elementary school children?. Arch Med Res, 2008; 39:796–802.

Boney

, Verma

, Tucker

, et al. Metabolic syndrome in childhood: Association with birth weight, maternal, and gestational diabetes mellitus. Pediatrics, 2005; 115:e290–e296.

Schellong

, Schulz

, Harder

, et al. Birth weight and long-term overweight risk: Systematic review and a meta-analysis including 643,902 persons from 66 studies and 26 countries globally. PLoS One, 2012; 7:e47776.

Hardy

, Mihrshahi

, Gale

, et al. 30-Year trends in overweight, obesity and waist-to-height ratio by socioeconomic status in Australian children, 1985 to 2015. Int J Obes (Lond), 2017; 41:76–82.

10.

www.censo2010.indec.gov.ar/. INDEC, 2010. Instituto Nacional de Estadística y Censos de la Republica Argentina (Accessed February, 2017).

11.

Tomeo

, Rich-Edwards

, Michels

, et al. Reproducibility and validity of maternal recall of pregnancy-related events. Epidemiology, 1999; 10:774–777.

12.

Kuczmarski

, Ogden

, Guo

, et al. 2000 CDC growth charts for the United States: Methods and development. Vital Health Stat, 2002; 11:1–190.

13.

Hirschler

, Oestreicher

, Maccallini

, et al. Relationship between obesity and metabolic syndrome among Argentinean elementary schoolchildren. Clin Biochem, 2010; 43:435–441.

14.

Flynn

, Kaelber

, Baker-Smith

, et al; Subcommittee on Screening and Management of High Blood Pressure in Children. Clinical practice guideline for screening and management of high blood pressure in children and adolescents. Pediatrics, 2017; 140:e20171904.

15.

Malina

, Katmarzyk

, Beunen

. Birth weight and its relationship to size attained and relative fat distribution at 7 to 12 years of age. Obes Res, 1996; 4:385–390.

16.

Rolfe

, Loos

, Druet

, et al. Association between birth weight and visceral fat in adults. Am J Clin Nutr, 2010; 92:347–352.

17.

Moore

, Harrall

, Sauder

, et al. Neonatal adiposity and childhood obesity. Pediatrics, 2020; 146:e20200737.

18.

Murtaugh

, Jacobs

Jr ., Moran

, et al. Relation of birth weight to fasting insulin, insulin resistance, and body size in adolescence. Diabetes Care, 2003; 26:187–192.

19.

Sousa

, Guimarães

, Daltro

, et al. Association between birth weight and cardiovascular risk factors in adolescents. Arq Bras Cardiol, 2013; 101:9–17.

20.

Rubin

. What's the best way to treat normal-weight people with metabolic abnormalities?. JAMA, 2018; 320:223–225.

21.

Javed

, Jumean

, Murad

, et al. Diagnostic performance of body mass index to identify obesity as defined by body adiposity in children and adolescents: A systematic review and meta-analysis. Pediatr Obes, 2015; 10:234–244.

22.

Wildman

, Muntner

, Reynolds

, et al. The obese without cardiometabolic risk factor clustering and the normal weight with cardiometabolic risk factor clustering: Prevalence and correlates of 2 phenotypes among the US population (NHANES 1999–2004). Arch Intern Med, 2008; 168:1617–1624.

23.

Ighbariya

, Weiss

. Insulin resistance, prediabetes, metabolic syndrome: What should every pediatrician know?. J Clin Res Pediatr Endocrinol, 2017; 9(Suppl 2):49–57.

24.

Sun

, Wang

, Heianza

, et al. Birthweight and cardiometabolic risk patterns in multiracial children. Int J Obes (Lond), 2018; 42:20–27.

25.

Singer

, Eng

, Lumeng

, et al. The relationship between body fat mass percentiles and inflammation in children. Obesity (Silver Spring), 2014; 22:1332–1336.

26.

Diels

, Vanden Berghe

, Van Hul

. Insights into the multifactorial causation of obesity by integrated genetic and epigenetic analysis. Obesity Rev, 2020; 21:e13019.