International Journal of Obesity Supplements (2015) 5, S74–S79 © 2015 Macmillan Publishers Limited All rights reserved 2046-2166/15 www.nature.com/ijosup
ORIGINAL ARTICLE
Birth weight and childhood obesity: a 12-country study Y Qiao1,2, J Ma1, Y Wang1, W Li1, PT Katzmarzyk2, J-P Chaput3, M Fogelholm4, WD Johnson2, R Kuriyan5, A Kurpad5, EV Lambert6, C Maher7, J Maia8, V Matsudo9, T Olds7, V Onywera10, OL Sarmiento11, M Standage12, MS Tremblay3, C Tudor-Locke2,13, TS Church2, P Zhao1 and G Hu2 for the ISCOLE Research Group OBJECTIVES: Few studies have investigated the association between the full range of birth weight and the risk of childhood obesity in high-, middle- and low-income countries. The aim of the present study is to assess the association between different levels of birth weight and the risk of obesity among children aged 9–11 years in 12 countries. METHODS: A multinational, cross-sectional study of 5141 children aged 9–11 years was conducted in 12 countries. Height and weight were obtained using standardized methods. Time spent in moderate-to-vigorous physical activity (MVPA), sedentary and sleeping were objectively measured using 24-h, waist-worn accelerometer (Actigraph GT3X+) monitored for 7 days. Birth weight and other factors (regions, parental education, maternal history of gestational diabetes, children age, gender, breast feeding, gestational age, unhealthy diet scores and healthy diet scores) were collected by parental and children’s questionnaires. Multilevel modeling was used to account for the nested nature of the data. RESULTS: The overall prevalence of obesity (BMI z-score4+2 s.d.) was 15.4% for boys and 10.0% for girls. There was a positive association between birth weight and BMI z-scores. The multivariable-adjusted odds ratios (ORs) of childhood obesity were significantly higher among children whose birth weights were 3500–3999 g (OR 1.45; 95% confidence interval (CI): 1.10–1.92), and 44000 g (OR 2.08; 95% CI: 1.47–2.93), compared with the reference group (2500–2999 g). The positive association between birth weight and the odds of childhood obesity was seen in girls, whereas a U-shaped association appeared in boys. CONCLUSIONS: High levels of birth weight, defined as birth weight ⩾ 3500 g, were associated with increased odds of obesity among 9–11-year-old children in 12 countries. However, sex differences in the association between birth weight and the risk of obesity need to be considered when planning interventions to reduce childhood obesity. International Journal of Obesity Supplements (2015) 5, S74–S79; doi:10.1038/ijosup.2015.23
INTRODUCTION Obesity is an important lifestyle-related public health problem worldwide.1 The prevalence of obesity in children has risen significantly during the past few decades not only in developed countries but also in developing countries.2 One recent review has reported that the prevalence of childhood overweight and obesity rose by 47.1% between 1980 and 2013 worldwide.3 At the same time, a rapidly increased rate of newborn macrosomia has been found in most developed and developing countries in the past two decades.4–6 Several studies have indicated that high birth weight is associated with an increased risk of childhood obesity.7–19 However, few studies have examined the extent to which birth weight is associated with obesity in young school children in high-, middle- and low-income countries. Moreover, the association between low birth weight and the risk of childhood obesity is controversial.7,8,10,12–18 Several studies have suggested that maternal and child factors, such as maternal history of gestational diabetes, infant feeding mode, moderate-to-vigorous physical activity (MVPA), diet, sedentary and sleeping times may confound the association between birth weight and later risk of childhood obesity,20–22 yet few studies were able to adjust for these factors
simultaneously. Therefore, the aim of the present study was to examine the association between birth weight and the risk of obesity in 9–11–year-old children from 12 countries, adjusting for several confounding variables.
MATERIALS AND METHODS Study design The International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE) is a multinational cross-sectional study conducted at urban and suburban sites in 12 countries (Australia, Brazil, Canada, China, Colombia, Finland, India, Kenya, Portugal, South Africa, the United Kingdom and the United States).23 These countries were divided into low- to high-income groups according to World Bank Classification (Table 1). More details on the study design and methods can be found elsewhere.23 The institutional review board at the Pennington Biomedical Research Center (coordinating center) approved the overarching protocol, and the institutional/ethical review boards at each participating institution also approved the local protocol. Written informed consent was obtained from parents or legal guardians, and child assent was also obtained as required by the local institutional/ethical review boards before participation in the study.
1 Tianjin Women’s and Children’s Health Center, Tianjin, China; 2Pennington Biomedical Research Center, Baton Rouge, LA, USA; 3Children’s Hospital of Eastern Ontario Research Institute, Ottawa, Ontario, Canada; 4Department of Food and Environmental Sciences, University of Helsinki, Helsinki, Finland; 5St Johns Research Institute, Bangalore, India; 6 Department of Human Biology, Faculty of Health Sciences, Division of Exercise Science and Sports Medicine, University of Cape Town, Cape Town, South Africa; 7Alliance for Research in Exercise Nutrition and Activity (ARENA), School of Health Sciences, University of South Australia, Adelaide, South Australia, Australia; 8CIFI2D, Faculdade de Desporto, University of Porto, Porto, Portugal; 9Centro de Estudos do Laboratório de Aptidão Física de São Caetano do Sul (CELAFISCS), Sao Paulo, Brazil; 10Department of Recreation Management and Exercise Science, Kenyatta University, Nairobi, Kenya; 11School of Medicine, Universidad de los Andes, Bogota, Colombia; 12University of Bath, Bath, UK and 13 Department of Kinesiology, University of Massachusetts Amherst, Amherst, MA, USA. Correspondence: Dr G Hu, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808, USA. E-mail: [email protected]
Birth weight and childhood obesity Y Qiao et al
S75 Table 1. International Study of Childhood Obesity, Lifestyle and the Environment field site characteristics Country
Australia Canada Finland Portugal UK USA Brazil China Colombia South Africa India Kenya Total
World bank classification
High income High income High income High income High income High income Upper-middle income Upper-middle income Upper-middle income Upper-middle income Lower-middle income Low income
Number of study samples Boys
Girls
Total
192 199 191 228 150 161 179 225 357 81 206 170 2339
219 273 223 318 196 226 193 202 358 137 249 208 2802
411 472 414 546 346 387 372 427 715 218 455 378 5141
Participants A total of 7372 children aged 9–11 years participated in ISCOLE, of which 5141 remained in the analytical sample for the present work after excluding participants who did not have valid data/information on accelerometer (n = 1214), birth weight (n = 698), body mass index (BMI; n = 5), gestational age (n = 121), diet scores (n = 85) or other information (highest parental education, maternal history of gestational diabetes, infant breast feeding; n = 108). Participants who were excluded from the present analysis did not differ in BMI z-scores, but the excluded sample had a higher proportion of boys than those who were included in the analysis. Data were collected from September 2011 through December 2013.
Measurements Demographics and family health history. A demographic and family health history questionnaire was completed by parents. The questionnaire collected information on highest parental education, maternal history of gestational diabetes, child age, sex, birth weight, infant feeding mode and gestational age. The highest parental education completed by the child’s mother was asked, as was the highest level of education completed by the father. These responses were collapsed into three categories: did not complete high school, completed high school or some college and completed bachelor or postgraduate degree. A variable was then created that represented the highest education level attained by either parent, and if one of the parents’ response to this question was missing, the highest level attained would be that of the other parent. Birth weight was reported by the child’s parent or guardian, and the data were divided into five categories: o2500 g, 2500–2999 g, 3000–3499 g, 3500–3999 g and ⩾ 4000 g. The child’s parent or guardian was asked whether the child was breast fed or not, the age when they completely stopped being fed breast milk, the age when they were first fed formula and the age when they completely stopped drinking formula. These responses were classified into four categories for the first 6 months: exclusive breast feeding, mixed feeding, weaned from breast feeding and exclusive formula feeding. Dietary intake. A Food Frequency Questionnaire that was adapted from the Health Behavior in School-aged Children Survey (HBSC)24 was administered to all ISCOLE student participants. The Food Frequency Questionnaire asks the participants about their ‘usual’ consumption of 23 food categories, with response categories including ‘never’, ‘less than once per week’, ‘once per week’, ‘2–4 days per week’, ‘5–6 days per week’, ‘once a day every day’ and ‘more than once a day’. Two diet scores which represented an ‘unhealthy diet pattern’ (with positive loadings for fast food, hamburgers, soft drinks, sweets, fried food, and so on) and a ‘healthy diet pattern’ (with positive loadings for vegetables, fruit, whole grains, lowfat milk, and so on) were obtained using principal components analyses.25 Anthropometry measurement. Height and weight were measured using standard procedures across all study sites. Height was measured without shoes using a Seca 213 portable stadiometer (Seca Corporation, © 2015 Macmillan Publishers Limited
Hamburg, Germany), with the participant’s head in the Frankfurt Plane. Weight was measured using a portable Tanita SC-240 Body Composition Analyzer (TANITA Corporation, Tokyo, Japan) after all outer clothing, heavy pocket items, shoes and socks were removed. Two measurements were obtained, and the average was used in analysis (a third measurement was obtained if the first two measurements were 40.5 cm or 0.5 kg apart for height and weight, respectively, and the average of the two closest measurements was used in analyses). BMI was calculated by dividing weight in kilograms by the square of height in meters. BMI z-scores were computed using age- and sex-specific reference data from the World Health Organization.26 Participants were classified as obese (BMI-for-age z-score4+2 s.d.) or non-obese (BMI-for-age z-score ⩽ +2 s.d.). Accelerometry. An ActiGraph GT3X+ accelerometer (ActiGraph, LLC, Pensacola, FL, USA) was used to objectively measure MVPA, sedentary behavior and sleeping time. The accelerometer was worn at the waist on an elasticized belt on the right mid-axillary line. Participants were encouraged to wear the accelerometer 24 h per day (removing only for water-related activities) for at least 7 days (plus an initial familiarization day and the morning of the final day), including 2 weekend days. The minimal amount of accelerometer data that was considered acceptable was 4 days with at least 10 h of waking wear time per day, including at least 1 weekend day.23,27 Nocturnal sleep duration was estimated from the accelerometer data using a fully automated algorithm for 24-h waist-worn accelerometers that was recently validated for ISCOLE.28,29 This new algorithm produces more precise estimates of sleep duration than previous algorithms and captures total sleep time from sleep onset to the end of sleep, including all epochs and wakefulness after onset.28 The weekly total sleep time averages were calculated using only days where valid sleep was accumulated (total sleep-period time ⩾ 160 min) and only for participants with at least 3 nights of valid sleep, including 1 weekend day.30 After exclusion of total sleep time and awake non-wear time (any sequence of ⩾ 20 consecutive minutes of 0 activity counts), MVPA was defined as all activity ⩾ 574 counts per 15 s and total SED as all movement ⩽ 25 counts per 15 s, consistent with the widely used Evenson cutoffs.29,31,32 Statistical analyses. Multilevel linear regression models were used to examine the association between birth weight and BMI-for-age z-scores. Multilevel logistic regression models were used to estimate the association between birth weight and the risk of childhood obesity. We defined child as level 1, school as level 2 and study site as level 3. Study site and school were considered to have random effects. As there was statistically significant interaction of birth weight and sex with the risk of childhood obesity, we analyzed separately by sex in some analyses. We used restricted cubic spline in logistic regression procedures to test whether there was a dose–response or non-linear association of birth weight as a continuous variable with the risk of childhood obesity. The analyses were adjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, MVPA, sleeping and sedentary times. The criterion for statistical significance was o 0.05. All statistical analyses were performed with SPSS for Windows, version 21.0 (Statistics 21, SPSS, IBM, Armonk, NY, USA) or SAS for Windows, version 9.4 (SAS Institute Inc., Cary, NC, USA).
RESULTS A total of 5141 children (2339 boys and 2802 girls) were included in the present study. The distribution of sample sizes across sites is presented in Table 1. The overall prevalence of obesity was 15.4% among boys and 10.0% among girls. The prevalence of low birth weight (o 2500 g) and macrosomia (⩾4000 g) was 6.8 and 9.5%, respectively. General characteristics of the study population are presented in Table 2. Table 3 compares the mean values of BMIfor-age z-scores at 9–11 years of age according to birth weight status categories. Children’s birth weight was positively associated with BMI-for-age z-scores among both girls and boys. The odds ratio (ORs) curves derived from spline logistic regression showed associations of the full range of birth weight with the odds of childhood obesity in boys and girls after adjustment for all confounding factors are presented in Figure 1. A nadir of the U-shaped association between birth weight and International Journal of Obesity Supplements (2015) S74 – S79
Birth weight and childhood obesity Y Qiao et al
S76 Table 2.
Characteristics of study participants Boys
Girls
Total
Age (years) 10.4 (0.56) 10.4 (0.56) 10.4 (0.56) Body weight (kg) 37.4 (9.4) 37.5 (9.5) 37.5 (9.4) Body height (cm) 141.7 (7.2) 142.2 (7.8) 141.9 (7.5) Body mass index (kg m − 2) 18.4 (3.4) 18.4 (3.4) 18.4 (3.4) Body mass index z-scoresa 0.6 (1.3) 0.4 (1.2) 0.5 (1.3) Unhealthy diet pattern score − 0.04 (0.96) − 0.18 (0.84) − 0.11 (0.90) Healthy diet pattern score − 0.05 (0.99) 0.04 (0.99) 0.00 (0.99) Duration of moderate-to-vigorous 69.1 (25.8) 52.2 (20.9) 59.9 (24.7) physical activity per day (min) Duration of sedentary per day (min) 508.3 (68.7) 524.2 (66.9) 516.9 (68.2) Duration of night sleep (min) 523.8 (52.8) 531.5 (53.0) 528.0 (53.0) Pregnancy term (weeks) 38.6 (2.1) 38.6 (2.2) 38.6 (2.2) Birth weight category, n (%) o 2500 g 2500–2999 g 3000–3499 g 3500–3999 g ⩾ 4000 g Body mass index category, n (%)b Non-obese Obese
128 444 777 708 282
(5.5) (19.0) (33.2) (30.3) (12.1)
222 606 1054 713 207
(7.9) (21.6) (37.6) (25.4) (7.4)
350 1050 1831 1421 489
(6.8) (20.4) (35.6) (27.6) (9.5)
1979 (84.6) 360 (15.4)
2523 (90.0) 279 (10.0)
4502 (87.6) 637 (12.4)
410 (17.5) 1013 (43.3)
496 (17.7) 1190 (42.5)
906 (17.6) 2203 (42.9)
916 (39.2)
1116 (39.8)
2032 (39.5)
Maternal history of gestational diabetes, n (%) No 2241 (95.8) Yes 98 (4.2)
2679 (95.6) 123 (4.4)
4920 (95.7) 221 (4.3)
Infant breast feeding, n (%) Exclusive breast feeding Mixed feeding Weaned from breast feeding Exclusive formula feeding
1023 1351 36 392
1923 2418 59 741
Parental education, n (%) Did not complete high school Completed high school/some college Bachelor’s degree or postgraduate degree
900 1067 23 349
(38.5) (45.6) (1.0) (14.9)
(36.5) (48.2) (1.3) (14.0)
(37.4) (47.0) (1.1) (14.4)
Data are means (s.d.) or number (percentage). aUsing age- and sex-specific reference data from the World Health Organization. bObesity was defined as BMI-for-age z-score4+2 s.d. and non-obesity was defined as BMI-for-age z-score ⩽ +2 s.d.
childhood obesity was observed among boys, and boys and girls combined whose birth weights were 2500–2999 g. Therefore, we used a birth weight range of 2500–2999 g as the reference in the categorical analyses. After adjustment for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, MVPA, sedentary behavior and sleep duration, the ORs of childhood obesity were significantly higher among children whose birth weights were 3500–3999 g (OR 1.45; 95% confidence interval (CI): 1.10–1.92) and ⩾ 4000 g (OR 2.08; 95% CI: 1.47–2.93), compared with the reference group (2500–2999 g; Table 4). There was a significant interaction between birth weight and sex with the odds of childhood obesity (P for interaction o 0.05). As shown in Figure 1a and b, there was a U-shaped association between birth weight and the odds of childhood obesity among boys, and a positive association between birth weight and the odds of childhood obesity among girls. After adjustment for all confounding factors, there were increased odds of childhood obesity among boys whose birth weights were ⩾ 4000 g (OR 1.77; 95% CI: 1.12–2.82) and among girls whose birth weights were 3500–3999 g (OR 1.66; 95% CI: 1.11–2.50) and ⩾ 4000 g (OR 2.48; 95% CI: 1.47–4.17), compared with the reference group (2500–2999 g). International Journal of Obesity Supplements (2015) S74 – S79
The association between birth weight and the odds of childhood obesity was also different between children from high-income countries and those from low/middle-income countries. ORs of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and started to rise after 3500 g of birth weight in children from low/ middle-income countries (Table 4). DISCUSSION This study demonstrated that higher levels of birth weight, defined as birth weight ⩾ 3500 g, were associated with an increased odd of obesity in 9–11-year-old children in 12 countries. The association between birth weight and the odds of childhood obesity had sex and country differences. The relation between birth weight and childhood obesity has been examined in many previous studies, yet none have had a sample consisting of children from all world regions. Most past studies have found a significant and positive association between birth weight and childhood obesity risk.7–19 A population-based cohort study from Denmark indicated an increased risk of overweight for children 6–13 years of age with birth weight ⩾ 4.0 kg compared with those with birth weight between 3.0 and 3.49 kg.12 Another Chinese birth cohort found that children with higher birth weight ⩾ 3500 g had an increased odds of childhood overweight at 3–6 years of age compared with those having birth weight of 3000–3249 g.15 A meta-analysis of 66 studies from 26 countries demonstrated that high birth weight (44000 g) was positively associated with increased odds of childhood overweight (OR 1.66; 95% CI: 1.55–1.77) compared with normal birth weight (2500–4000 g).8 However, most of these studies only assessed high birth weight defined as birth weight 44000 g but not the full range of birth weight with the risk of childhood obesity. One Chinese study examined the association of the full range of birth weight with the risk of childhood obesity at 3–6 years of age, and found that the ORs of becoming overweight plus obese during childhood increased significantly when birth weights were ⩾ 3000 g, and becoming obese during childhood increased significantly when birth weights were ⩾ 3500 g compared with birth weight 2500–2999 g,14 which is the same as our study. Several studies have indicated that low birth weight is an important risk factor for adult disease occurrence, especially for hypertension, metabolic disorders and other chronic diseases later in life.33,34 Intrauterine retardation may progress risk factors for development of metabolic disease during adulthood. Increased risks are observed particularly in those with a low birth weight and being overweight or obese, and physically inactive during adulthood. Some studies have assessed the association between low birth weight and the risk of childhood obesity, and the results are controversial7,8,10,12–18 Several studies have found that low birth weight was associated with a reduced risk of childhood obesity.15,17–19 However, other studies have indicated no association,7,12,14 or an increased risk of obesity with low birth weight.10,13 The present study was underpowered to identify any association between low birth weight and children obesity. Small sample sizes and low prevalence of obesity in children with low birth weight may limit the statistical power in the analyses of associations between low birth weight and the risk of childhood obesity. These features directly address some of the primary reasons why earlier studies may have produced contrasting results. A recent Chinese study including very large samples (1703 children with low birth weight and 55 925 children in total) assessed the association of birth weight with the risk of obesity in children aged o 3 years.18 Results showed a positive association between birth weight and childhood obesity, indicating that low levels of birth weight, defined as birth weight o 2500 g, were associated with a decreased risk of childhood obesity, and higher © 2015 Macmillan Publishers Limited
Birth weight and childhood obesity Y Qiao et al
S77 Table 3.
Mean difference of BMI z-scores in 9–11–year-old children based on birth weight status P-value for trend
Birth weight (g) o 2500
2500–2999
3000–3499
3500–3999
⩾ 4000
Total Number of participants BMI z-score Multiple-adjusted BMI z-scoresb
350 − 0.05 (0.07) − 0.11 (0.08)
1050 0 0
1831 0.10 (0.05) 0.10 (0.05)
1421 0.26 (0.05) 0.25 (0.05)
489 0.44 (0.07) 0.47 (0.07)
o0.001 o0.001
Boys Number of participants BMI-for-age z-score Multiple-adjusted BMI z-scoreb
128 0.03 (0.13) − 0.04 (0.13)
444 0 0
777 0.05 (0.08) 0.06 (0.07)
708 0.17 (0.08) 0.15 (0.08)
282 0.39 (0.10) 0.43 (0.10)
o0.001 o0.001
Girls Number of participants BMI-for-age z-score Multiple-adjusted BMI z-scoreb
222 − 0.10 (0.09) −0.16 (0.10)
606 0 0
1054 0.14 (0.06) 0.14 (0.06)
713 0.32 (0.07) 0.34 (0.07)
207 0.47 (0.10) 0.50 (0.10)
o0.001 o0.001
a
Abbreviation: BMI, body mass index. Data are mean difference (s.e.m.). aAdjusted also for sex. bAdjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times.
Figure 1. OR of continuous birth weight in relation to its risk for obesity among boys (a), girls (b), and boys and girls combined (c). Adjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times. Lines with short dashes represent the pointwise 95% CIs. © 2015 Macmillan Publishers Limited
International Journal of Obesity Supplements (2015) S74 – S79
Birth weight and childhood obesity Y Qiao et al
S78 Table 4.
Odds ratio for obesity in 9–11-year-old children based on birth weight status P-value for trend
Birth weight (g) o 2500
2500–2999
3000–3499
3500–3999
⩾ 4000
Total Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
350 34 1.05 (0.69–1.59) 1.02 (0.65–1.59)
1050 101 1 1
1831 217 1.18 (0.91–1.52) 1.18 (0.91–1.54)
1421 201 1.45 (1.12–1.89) 1.45 (1.10–1.92)
489 86 1.91 (1.38–2.64) 2.08 (1.47–2.93)
o0.001 o0.001
Boys Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
128 21 1.38 (0.79–2.43) 1.37 (0.75–2.50)
444 55 1 1
777 114 1.08 (0.76–1.55) 1.08 (0.74–1.57)
708 118 1.34 (0.94–1.93) 1.28 (0.88–1.88)
282 52 1.59 (1.03–2.47) 1.77 (1.12–2.82)
0.154 0.093
Girls Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
222 13 0.74 (0.39–1.40) 0.71 (0.36–1.41)
606 46 1 1
1054 103 1.25 (0.86–1.80) 1.28 (0.88–1.88)
713 83 1.55 (1.06–2.29) 1.67 (1.11–2.50)
207 34 2.37 (1.45–3.86) 2.48 (1.47–4.17)
0.001 0.002
High-income countriesa Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
165 20 1.08 (0.61–1.91) 1.04 (0.56–1.94)
407 46 1 1
920 99 1.01 (0.69–1.47) 0.97 (0.65–1.43)
773 91 1.20 (0.81–1.76) 1.22 (0.81–1.83)
311 48 1.70 (1.09–2.66) 1.79 (1.11–2.89)
0.087 0.045
Low/middle-income countriesa Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
185 14 0.94 (0.50–1.76) 0.91 (0.47–1.76)
643 55 1 1
911 118 1.32 (0.93–1.87) 1.33 (0.92–1.92)
648 110 1.68 (1.17–2.41) 1.61 (1.10–2.35)
178 38 2.07 (1.28–3.35) 2.30 (1.38–3.83)
o0.001 0.011
a
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio. aAdjusted also for sex. bAdjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times.
levels of birth weight, defined as birth weight 43000 g, were associated with an increased risk of childhood obesity. To date, four studies have evaluated whether there is a sexspecific association between birth weight and the risk of childhood obesity.12,14,16,18 All four studies have found a positive association between birth weight and the risk of childhood obesity in girls. However, the associations between birth weight and the risk of childhood obesity in boys were less consistent. Three studies described a positive association between birth weight and the risk of childhood obesity,12,16,18 whereas one Chinese study found a U-shaped association between birth weight and the risk of childhood obesity.14 The research with Chinese children showed low levels of birth weight not to be associated with increased risk of childhood obesity in boys (OR 2.29; 95% CI: 0.80–6.50) or decreased risk of childhood obesity in girls (OR 0.66; 95% CI: 0.08–5.33).14 Such findings are consistent with those reported in the present study. Thus, more studies are needed to assess whether there is a sex difference in the association between birth weight and childhood obesity. The present study also indicated that the extent of high birth weight with the risk of childhood obesity was different in different countries. The relative risk of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and started to rise after 3500 g of birth weight in children from low/middle-income countries. A recent study found that higher levels of birth weight, defined as birth weight 43000 g, were associated with an increased risk of overweight or obesity among Chinese children from 6 months to 3 years of age (middle-income countries).18 Several studies in developed countries have shown higher levels of birth weight, defined as birth weight 44000 g, to International Journal of Obesity Supplements (2015) S74 – S79
be positively associated with an increased risk of childhood obesity.13,16 All these results seem to support our finding. The differences in maternal weight and nutritional status before and during pregnancy and child feeding between developed and developing countries might result in the different extent of association between high birth weight and the risk of childhood obesity between these countries. There are several strengths in our study including the recruitment of a large multinational sample of children from low- to high-income countries across several regions of the world, the highly standardized measurement protocol, the use of direct measurements whenever possible and the rigorous quality-control program. In addition, MVPA, sedentary behavior and sleep duration were objectively obtained from the 24-h, waist-worn accelerometer, and dietary intake was assessed by Food Frequency Questionnaire in the present study. There are also a number of limitations to the present study. First, the crosssectional study precludes us from making cause-and-effect inferences. Second, as the information on maternal prepregnancy BMI were not available in the present study, we may not be able to fully control for the effect of this variable on the association of birth weight with the risk of childhood obesity. Third, the birth weight records may have been recalled wrongly by the parent or guardian. The degree to which these factors may have biased the results is unknown. In summary, higher levels of birth weight, defined as birth weight ⩾ 3500 g, were positively associated with an increased risk of obesity in 9–11-year-old children in 12 countries. There was a U-shaped association between birth weight and the risk of childhood obesity among boys, and a positive linear association © 2015 Macmillan Publishers Limited
Birth weight and childhood obesity Y Qiao et al
S79 between birth weight and the risk of childhood obesity among girls. The relative risks of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and after 3500 g of birth weight in children from low/middleincome countries. CONFLICT OF INTEREST MF has received a research grant from Fazer Finland and has received an honorarium for speaking for Merck. AK has been a member of the Advisory Boards of Dupont and McCain Foods. RK has received a research grant from Abbott Nutrition Research and Development. VM is a member of the Scientific Advisory Board of Actigraph and has received an honorarium for speaking for The Coca-Cola Company. TO has received an honorarium for speaking for The Coca-Cola Company. The remaining authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank the ISCOLE External Advisory Board and the ISCOLE participants and their families who made this study possible. A membership list of the ISCOLE Research Group and External Advisory Board is included in Katzmarzyk et al. (this issue). ISCOLE was funded by The Coca-Cola Company.
REFERENCES 1 World Health Organization. Obesity: preventing and managing the global epidemic. Geneva: World Health Organization; 2000. World Health Organ Tech Rep Ser 2013; 894: 1–252. 2 Wu JF. Childhood obesity: a growing global health hazard extending to adulthood. Pediatr Neonatol 2013; 54: 71–72. 3 Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2014; 384: 766–781. 4 Lu Y, Zhang J, Lu X, Xi W, Li Z. Secular trends of macrosomia in southeast China, 1994-2005. BMC Public Health 2011; 11: 818–818. 5 Koyanagi A, Zhang J, Dagvadorj A, Hirayama F, Shibuya K, Souza JP et al. Macrosomia in 23 developing countries: an analysis of a multicountry, facility-based, cross-sectional survey. Lancet 2013; 381: 476–483. 6 Schack-Nielsen L, Molgaard C, Sorensen TI, Greisen G, Michaelsen KF. Secular change in size at birth from 1973 to 2003: national data from Denmark. Obesity 2006; 14: 1257–1263. 7 Yu ZB, Han SP, Zhu GZ, Zhu C, Wang XJ, Cao XG et al. Birth weight and subsequent risk of obesity: a systematic review and meta-analysis. Obes Rev 2011; 12: 525–542. 8 Schellong K, Schulz S, Harder T, Plagemann A. 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. 9 Weng SF, Redsell SA, Swift JA, Yang M, Glazebrook CP. Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy. Arch Dis Child 2012; 97: 1019–1026. 10 Frye C, Heinrich J. Trends and predictors of overweight and obesity in East German children. Int J Obes Relat Metab Disord 2003; 27: 963–969. 11 Monteiro PO, Victora CG, Barros FC, Monteiro LM. Birth size, early childhood growth, and adolescent obesity in a Brazilian birth cohort. Int J Obes Relat Metab Disord 2003; 27: 1274–1282. 12 Rugholm S, Baker JL, Olsen LW, Schack-Nielsen L, Bua J, Sorensen TI. Stability of the association between birth weight and childhood overweight during the development of the obesity epidemic. Obes Res 2005; 13: 2187–2194. 13 Dubois L, Girard M. Early determinants of overweight at 4.5 years in a populationbased longitudinal study. Int J Obes 2006; 30: 610–617.
© 2015 Macmillan Publishers Limited
14 Zhang X, Liu E, Tian Z, Wang W, Ye T, Liu G et al. High birth weight and overweight or obesity among Chinese children 3-6 years old. Prev Med 2009; 49: 172–178. 15 Ye R, Pei L, Ren A, Zhang Y, Zheng X, Liu J-m. Birth weight, maternal body mass index, and early childhood growth: a prospective birth cohort study in China. J Epidemiol 2010; 20: 421–428. 16 Oldroyd J, Renzaho A, Skouteris H. Low and high birth weight as risk factors for obesity among 4 to 5-year-old Australian children: does gender matter? Eur J Pediatr 2011; 170: 899–906. 17 Loaiza S, Coustasse A, Urrutia-Rojas X, Atalah E. Birth weight and obesity risk at first grade in a cohort of Chilean children. Nutr Hosp 2011; 26: 214–219. 18 Li N, Liu E, Sun S, Guo J, Pan L, Wang P et al. Birth weight and overweight or obesity risk in children under 3 years in China. Am J Hum Biol 2014; 26: 331–336. 19 Rooney BL, Mathiason MA, Schauberger CW. Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort. Matern Child Health J 2011; 15: 1166–1175. 20 Hart CN, Cairns A, Jelalian E. Sleep and obesity in children and adolescents. Pediatr Clin North Am 2011; 58: 715–733. 21 Pearson N, Biddle SJ. Sedentary behavior and dietary intake in children, adolescents, and adults. A systematic review. Am J Prev Med 2011; 41: 178–188. 22 Prentice-Dunn H, Prentice-Dunn S. Physical activity, sedentary behavior, and childhood obesity: a review of cross-sectional studies. Psychol Health Med 2012; 17: 255–273. 23 Katzmarzyk PT, Barreira TV, Broyles ST, Champagne CM, Chaput JP, Fogelholm M et al. The International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE): design and methods. BMC Public Health 2013; 13: 900. 24 Currie C, Gabhainn SN, Godeau E, Roberts C, Smith R, Currie D et al. (eds) Inequalities in Children's Health: HBSC International Report from the 2005/2006 Survey World Health Organization: Geneva, Switzerland, 2008. 25 Mikkilä V, Vepsäläinen H, Saloheimo T, Gonzalez SA, Meisel JD, Hu G et al. An international comparison of dietary patterns in 9-11-year old children. Int J Obes Suppl 2015 (this issue). 26 De Onis M, Onyanga AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull WHO 2007; 85: 660–667. 27 Tudor-Locke C, Mire EF, Dentro KN, Barreira TV, Schuna JM, Zhao P et al. A model for presenting accelerometer paradata in large studies: ISCOLE. Int J Behav Nutr Phys Act 2015; 12: 52. 28 Barreira TV, Schuna Jr JM, Mire EF, Katzmarzyk PT, Chaput JP, Leduc G et al. Identifying children's nocturnal sleep using 24-h waist accelerometry. Med Sci Sports Exerc 2015; 47: 937–943. 29 Tudor-Locke C, Barreira TV, Schuna JM, Mire EF, Chaput J-P, Fogelholm M et al. Improving wear time compliance with a 24-hour waist-worn accelerometer protocol in the International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE). Int J Behav Nutr Phys Act 2015; 12: 11. 30 Tudor-Locke C, Barreira TV, Schuna Jr JM, Mire EF, Katzmarzyk PT. Fully automated waist-worn accelerometer algorithm for detecting children's sleep-period time separate from 24-h physical activity or sedentary behaviors. Appl Physiol Nutr Me 2014; 39: 53–57. 31 Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sport Sci 2008; 26: 1557–1565. 32 Katzmarzyk PT, Barreira TV, Broyles ST, Champagne CM, Chaput J-P, Fogelholm M et al. Physical activity, sedentary time, and obesity in an international sample of children. Med Sci Sports Exerc 2015; 47: 2062–2069. 33 Eriksson JG. Early growth and adult health outcomes–lessons learned from the Helsinki Birth Cohort Study. Matern Child Nutr 2005; 1: 149–154. 34 Chen W, Srinivasan SR, Yao L, Li S, Dasmahapatra P, Fernandez C et al. Low birth weight is associated with higher blood pressure variability from childhood to young adulthood: the Bogalusa Heart Study. Am J Epidemiol 2012; 176 (Suppl 7): S99–S105.
International Journal of Obesity Supplements (2015) S74 – S79
ORIGINAL ARTICLE
Birth weight and childhood obesity: a 12-country study Y Qiao1,2, J Ma1, Y Wang1, W Li1, PT Katzmarzyk2, J-P Chaput3, M Fogelholm4, WD Johnson2, R Kuriyan5, A Kurpad5, EV Lambert6, C Maher7, J Maia8, V Matsudo9, T Olds7, V Onywera10, OL Sarmiento11, M Standage12, MS Tremblay3, C Tudor-Locke2,13, TS Church2, P Zhao1 and G Hu2 for the ISCOLE Research Group OBJECTIVES: Few studies have investigated the association between the full range of birth weight and the risk of childhood obesity in high-, middle- and low-income countries. The aim of the present study is to assess the association between different levels of birth weight and the risk of obesity among children aged 9–11 years in 12 countries. METHODS: A multinational, cross-sectional study of 5141 children aged 9–11 years was conducted in 12 countries. Height and weight were obtained using standardized methods. Time spent in moderate-to-vigorous physical activity (MVPA), sedentary and sleeping were objectively measured using 24-h, waist-worn accelerometer (Actigraph GT3X+) monitored for 7 days. Birth weight and other factors (regions, parental education, maternal history of gestational diabetes, children age, gender, breast feeding, gestational age, unhealthy diet scores and healthy diet scores) were collected by parental and children’s questionnaires. Multilevel modeling was used to account for the nested nature of the data. RESULTS: The overall prevalence of obesity (BMI z-score4+2 s.d.) was 15.4% for boys and 10.0% for girls. There was a positive association between birth weight and BMI z-scores. The multivariable-adjusted odds ratios (ORs) of childhood obesity were significantly higher among children whose birth weights were 3500–3999 g (OR 1.45; 95% confidence interval (CI): 1.10–1.92), and 44000 g (OR 2.08; 95% CI: 1.47–2.93), compared with the reference group (2500–2999 g). The positive association between birth weight and the odds of childhood obesity was seen in girls, whereas a U-shaped association appeared in boys. CONCLUSIONS: High levels of birth weight, defined as birth weight ⩾ 3500 g, were associated with increased odds of obesity among 9–11-year-old children in 12 countries. However, sex differences in the association between birth weight and the risk of obesity need to be considered when planning interventions to reduce childhood obesity. International Journal of Obesity Supplements (2015) 5, S74–S79; doi:10.1038/ijosup.2015.23
INTRODUCTION Obesity is an important lifestyle-related public health problem worldwide.1 The prevalence of obesity in children has risen significantly during the past few decades not only in developed countries but also in developing countries.2 One recent review has reported that the prevalence of childhood overweight and obesity rose by 47.1% between 1980 and 2013 worldwide.3 At the same time, a rapidly increased rate of newborn macrosomia has been found in most developed and developing countries in the past two decades.4–6 Several studies have indicated that high birth weight is associated with an increased risk of childhood obesity.7–19 However, few studies have examined the extent to which birth weight is associated with obesity in young school children in high-, middle- and low-income countries. Moreover, the association between low birth weight and the risk of childhood obesity is controversial.7,8,10,12–18 Several studies have suggested that maternal and child factors, such as maternal history of gestational diabetes, infant feeding mode, moderate-to-vigorous physical activity (MVPA), diet, sedentary and sleeping times may confound the association between birth weight and later risk of childhood obesity,20–22 yet few studies were able to adjust for these factors
simultaneously. Therefore, the aim of the present study was to examine the association between birth weight and the risk of obesity in 9–11–year-old children from 12 countries, adjusting for several confounding variables.
MATERIALS AND METHODS Study design The International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE) is a multinational cross-sectional study conducted at urban and suburban sites in 12 countries (Australia, Brazil, Canada, China, Colombia, Finland, India, Kenya, Portugal, South Africa, the United Kingdom and the United States).23 These countries were divided into low- to high-income groups according to World Bank Classification (Table 1). More details on the study design and methods can be found elsewhere.23 The institutional review board at the Pennington Biomedical Research Center (coordinating center) approved the overarching protocol, and the institutional/ethical review boards at each participating institution also approved the local protocol. Written informed consent was obtained from parents or legal guardians, and child assent was also obtained as required by the local institutional/ethical review boards before participation in the study.
1 Tianjin Women’s and Children’s Health Center, Tianjin, China; 2Pennington Biomedical Research Center, Baton Rouge, LA, USA; 3Children’s Hospital of Eastern Ontario Research Institute, Ottawa, Ontario, Canada; 4Department of Food and Environmental Sciences, University of Helsinki, Helsinki, Finland; 5St Johns Research Institute, Bangalore, India; 6 Department of Human Biology, Faculty of Health Sciences, Division of Exercise Science and Sports Medicine, University of Cape Town, Cape Town, South Africa; 7Alliance for Research in Exercise Nutrition and Activity (ARENA), School of Health Sciences, University of South Australia, Adelaide, South Australia, Australia; 8CIFI2D, Faculdade de Desporto, University of Porto, Porto, Portugal; 9Centro de Estudos do Laboratório de Aptidão Física de São Caetano do Sul (CELAFISCS), Sao Paulo, Brazil; 10Department of Recreation Management and Exercise Science, Kenyatta University, Nairobi, Kenya; 11School of Medicine, Universidad de los Andes, Bogota, Colombia; 12University of Bath, Bath, UK and 13 Department of Kinesiology, University of Massachusetts Amherst, Amherst, MA, USA. Correspondence: Dr G Hu, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808, USA. E-mail: [email protected]
Birth weight and childhood obesity Y Qiao et al
S75 Table 1. International Study of Childhood Obesity, Lifestyle and the Environment field site characteristics Country
Australia Canada Finland Portugal UK USA Brazil China Colombia South Africa India Kenya Total
World bank classification
High income High income High income High income High income High income Upper-middle income Upper-middle income Upper-middle income Upper-middle income Lower-middle income Low income
Number of study samples Boys
Girls
Total
192 199 191 228 150 161 179 225 357 81 206 170 2339
219 273 223 318 196 226 193 202 358 137 249 208 2802
411 472 414 546 346 387 372 427 715 218 455 378 5141
Participants A total of 7372 children aged 9–11 years participated in ISCOLE, of which 5141 remained in the analytical sample for the present work after excluding participants who did not have valid data/information on accelerometer (n = 1214), birth weight (n = 698), body mass index (BMI; n = 5), gestational age (n = 121), diet scores (n = 85) or other information (highest parental education, maternal history of gestational diabetes, infant breast feeding; n = 108). Participants who were excluded from the present analysis did not differ in BMI z-scores, but the excluded sample had a higher proportion of boys than those who were included in the analysis. Data were collected from September 2011 through December 2013.
Measurements Demographics and family health history. A demographic and family health history questionnaire was completed by parents. The questionnaire collected information on highest parental education, maternal history of gestational diabetes, child age, sex, birth weight, infant feeding mode and gestational age. The highest parental education completed by the child’s mother was asked, as was the highest level of education completed by the father. These responses were collapsed into three categories: did not complete high school, completed high school or some college and completed bachelor or postgraduate degree. A variable was then created that represented the highest education level attained by either parent, and if one of the parents’ response to this question was missing, the highest level attained would be that of the other parent. Birth weight was reported by the child’s parent or guardian, and the data were divided into five categories: o2500 g, 2500–2999 g, 3000–3499 g, 3500–3999 g and ⩾ 4000 g. The child’s parent or guardian was asked whether the child was breast fed or not, the age when they completely stopped being fed breast milk, the age when they were first fed formula and the age when they completely stopped drinking formula. These responses were classified into four categories for the first 6 months: exclusive breast feeding, mixed feeding, weaned from breast feeding and exclusive formula feeding. Dietary intake. A Food Frequency Questionnaire that was adapted from the Health Behavior in School-aged Children Survey (HBSC)24 was administered to all ISCOLE student participants. The Food Frequency Questionnaire asks the participants about their ‘usual’ consumption of 23 food categories, with response categories including ‘never’, ‘less than once per week’, ‘once per week’, ‘2–4 days per week’, ‘5–6 days per week’, ‘once a day every day’ and ‘more than once a day’. Two diet scores which represented an ‘unhealthy diet pattern’ (with positive loadings for fast food, hamburgers, soft drinks, sweets, fried food, and so on) and a ‘healthy diet pattern’ (with positive loadings for vegetables, fruit, whole grains, lowfat milk, and so on) were obtained using principal components analyses.25 Anthropometry measurement. Height and weight were measured using standard procedures across all study sites. Height was measured without shoes using a Seca 213 portable stadiometer (Seca Corporation, © 2015 Macmillan Publishers Limited
Hamburg, Germany), with the participant’s head in the Frankfurt Plane. Weight was measured using a portable Tanita SC-240 Body Composition Analyzer (TANITA Corporation, Tokyo, Japan) after all outer clothing, heavy pocket items, shoes and socks were removed. Two measurements were obtained, and the average was used in analysis (a third measurement was obtained if the first two measurements were 40.5 cm or 0.5 kg apart for height and weight, respectively, and the average of the two closest measurements was used in analyses). BMI was calculated by dividing weight in kilograms by the square of height in meters. BMI z-scores were computed using age- and sex-specific reference data from the World Health Organization.26 Participants were classified as obese (BMI-for-age z-score4+2 s.d.) or non-obese (BMI-for-age z-score ⩽ +2 s.d.). Accelerometry. An ActiGraph GT3X+ accelerometer (ActiGraph, LLC, Pensacola, FL, USA) was used to objectively measure MVPA, sedentary behavior and sleeping time. The accelerometer was worn at the waist on an elasticized belt on the right mid-axillary line. Participants were encouraged to wear the accelerometer 24 h per day (removing only for water-related activities) for at least 7 days (plus an initial familiarization day and the morning of the final day), including 2 weekend days. The minimal amount of accelerometer data that was considered acceptable was 4 days with at least 10 h of waking wear time per day, including at least 1 weekend day.23,27 Nocturnal sleep duration was estimated from the accelerometer data using a fully automated algorithm for 24-h waist-worn accelerometers that was recently validated for ISCOLE.28,29 This new algorithm produces more precise estimates of sleep duration than previous algorithms and captures total sleep time from sleep onset to the end of sleep, including all epochs and wakefulness after onset.28 The weekly total sleep time averages were calculated using only days where valid sleep was accumulated (total sleep-period time ⩾ 160 min) and only for participants with at least 3 nights of valid sleep, including 1 weekend day.30 After exclusion of total sleep time and awake non-wear time (any sequence of ⩾ 20 consecutive minutes of 0 activity counts), MVPA was defined as all activity ⩾ 574 counts per 15 s and total SED as all movement ⩽ 25 counts per 15 s, consistent with the widely used Evenson cutoffs.29,31,32 Statistical analyses. Multilevel linear regression models were used to examine the association between birth weight and BMI-for-age z-scores. Multilevel logistic regression models were used to estimate the association between birth weight and the risk of childhood obesity. We defined child as level 1, school as level 2 and study site as level 3. Study site and school were considered to have random effects. As there was statistically significant interaction of birth weight and sex with the risk of childhood obesity, we analyzed separately by sex in some analyses. We used restricted cubic spline in logistic regression procedures to test whether there was a dose–response or non-linear association of birth weight as a continuous variable with the risk of childhood obesity. The analyses were adjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, MVPA, sleeping and sedentary times. The criterion for statistical significance was o 0.05. All statistical analyses were performed with SPSS for Windows, version 21.0 (Statistics 21, SPSS, IBM, Armonk, NY, USA) or SAS for Windows, version 9.4 (SAS Institute Inc., Cary, NC, USA).
RESULTS A total of 5141 children (2339 boys and 2802 girls) were included in the present study. The distribution of sample sizes across sites is presented in Table 1. The overall prevalence of obesity was 15.4% among boys and 10.0% among girls. The prevalence of low birth weight (o 2500 g) and macrosomia (⩾4000 g) was 6.8 and 9.5%, respectively. General characteristics of the study population are presented in Table 2. Table 3 compares the mean values of BMIfor-age z-scores at 9–11 years of age according to birth weight status categories. Children’s birth weight was positively associated with BMI-for-age z-scores among both girls and boys. The odds ratio (ORs) curves derived from spline logistic regression showed associations of the full range of birth weight with the odds of childhood obesity in boys and girls after adjustment for all confounding factors are presented in Figure 1. A nadir of the U-shaped association between birth weight and International Journal of Obesity Supplements (2015) S74 – S79
Birth weight and childhood obesity Y Qiao et al
S76 Table 2.
Characteristics of study participants Boys
Girls
Total
Age (years) 10.4 (0.56) 10.4 (0.56) 10.4 (0.56) Body weight (kg) 37.4 (9.4) 37.5 (9.5) 37.5 (9.4) Body height (cm) 141.7 (7.2) 142.2 (7.8) 141.9 (7.5) Body mass index (kg m − 2) 18.4 (3.4) 18.4 (3.4) 18.4 (3.4) Body mass index z-scoresa 0.6 (1.3) 0.4 (1.2) 0.5 (1.3) Unhealthy diet pattern score − 0.04 (0.96) − 0.18 (0.84) − 0.11 (0.90) Healthy diet pattern score − 0.05 (0.99) 0.04 (0.99) 0.00 (0.99) Duration of moderate-to-vigorous 69.1 (25.8) 52.2 (20.9) 59.9 (24.7) physical activity per day (min) Duration of sedentary per day (min) 508.3 (68.7) 524.2 (66.9) 516.9 (68.2) Duration of night sleep (min) 523.8 (52.8) 531.5 (53.0) 528.0 (53.0) Pregnancy term (weeks) 38.6 (2.1) 38.6 (2.2) 38.6 (2.2) Birth weight category, n (%) o 2500 g 2500–2999 g 3000–3499 g 3500–3999 g ⩾ 4000 g Body mass index category, n (%)b Non-obese Obese
128 444 777 708 282
(5.5) (19.0) (33.2) (30.3) (12.1)
222 606 1054 713 207
(7.9) (21.6) (37.6) (25.4) (7.4)
350 1050 1831 1421 489
(6.8) (20.4) (35.6) (27.6) (9.5)
1979 (84.6) 360 (15.4)
2523 (90.0) 279 (10.0)
4502 (87.6) 637 (12.4)
410 (17.5) 1013 (43.3)
496 (17.7) 1190 (42.5)
906 (17.6) 2203 (42.9)
916 (39.2)
1116 (39.8)
2032 (39.5)
Maternal history of gestational diabetes, n (%) No 2241 (95.8) Yes 98 (4.2)
2679 (95.6) 123 (4.4)
4920 (95.7) 221 (4.3)
Infant breast feeding, n (%) Exclusive breast feeding Mixed feeding Weaned from breast feeding Exclusive formula feeding
1023 1351 36 392
1923 2418 59 741
Parental education, n (%) Did not complete high school Completed high school/some college Bachelor’s degree or postgraduate degree
900 1067 23 349
(38.5) (45.6) (1.0) (14.9)
(36.5) (48.2) (1.3) (14.0)
(37.4) (47.0) (1.1) (14.4)
Data are means (s.d.) or number (percentage). aUsing age- and sex-specific reference data from the World Health Organization. bObesity was defined as BMI-for-age z-score4+2 s.d. and non-obesity was defined as BMI-for-age z-score ⩽ +2 s.d.
childhood obesity was observed among boys, and boys and girls combined whose birth weights were 2500–2999 g. Therefore, we used a birth weight range of 2500–2999 g as the reference in the categorical analyses. After adjustment for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, MVPA, sedentary behavior and sleep duration, the ORs of childhood obesity were significantly higher among children whose birth weights were 3500–3999 g (OR 1.45; 95% confidence interval (CI): 1.10–1.92) and ⩾ 4000 g (OR 2.08; 95% CI: 1.47–2.93), compared with the reference group (2500–2999 g; Table 4). There was a significant interaction between birth weight and sex with the odds of childhood obesity (P for interaction o 0.05). As shown in Figure 1a and b, there was a U-shaped association between birth weight and the odds of childhood obesity among boys, and a positive association between birth weight and the odds of childhood obesity among girls. After adjustment for all confounding factors, there were increased odds of childhood obesity among boys whose birth weights were ⩾ 4000 g (OR 1.77; 95% CI: 1.12–2.82) and among girls whose birth weights were 3500–3999 g (OR 1.66; 95% CI: 1.11–2.50) and ⩾ 4000 g (OR 2.48; 95% CI: 1.47–4.17), compared with the reference group (2500–2999 g). International Journal of Obesity Supplements (2015) S74 – S79
The association between birth weight and the odds of childhood obesity was also different between children from high-income countries and those from low/middle-income countries. ORs of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and started to rise after 3500 g of birth weight in children from low/ middle-income countries (Table 4). DISCUSSION This study demonstrated that higher levels of birth weight, defined as birth weight ⩾ 3500 g, were associated with an increased odd of obesity in 9–11-year-old children in 12 countries. The association between birth weight and the odds of childhood obesity had sex and country differences. The relation between birth weight and childhood obesity has been examined in many previous studies, yet none have had a sample consisting of children from all world regions. Most past studies have found a significant and positive association between birth weight and childhood obesity risk.7–19 A population-based cohort study from Denmark indicated an increased risk of overweight for children 6–13 years of age with birth weight ⩾ 4.0 kg compared with those with birth weight between 3.0 and 3.49 kg.12 Another Chinese birth cohort found that children with higher birth weight ⩾ 3500 g had an increased odds of childhood overweight at 3–6 years of age compared with those having birth weight of 3000–3249 g.15 A meta-analysis of 66 studies from 26 countries demonstrated that high birth weight (44000 g) was positively associated with increased odds of childhood overweight (OR 1.66; 95% CI: 1.55–1.77) compared with normal birth weight (2500–4000 g).8 However, most of these studies only assessed high birth weight defined as birth weight 44000 g but not the full range of birth weight with the risk of childhood obesity. One Chinese study examined the association of the full range of birth weight with the risk of childhood obesity at 3–6 years of age, and found that the ORs of becoming overweight plus obese during childhood increased significantly when birth weights were ⩾ 3000 g, and becoming obese during childhood increased significantly when birth weights were ⩾ 3500 g compared with birth weight 2500–2999 g,14 which is the same as our study. Several studies have indicated that low birth weight is an important risk factor for adult disease occurrence, especially for hypertension, metabolic disorders and other chronic diseases later in life.33,34 Intrauterine retardation may progress risk factors for development of metabolic disease during adulthood. Increased risks are observed particularly in those with a low birth weight and being overweight or obese, and physically inactive during adulthood. Some studies have assessed the association between low birth weight and the risk of childhood obesity, and the results are controversial7,8,10,12–18 Several studies have found that low birth weight was associated with a reduced risk of childhood obesity.15,17–19 However, other studies have indicated no association,7,12,14 or an increased risk of obesity with low birth weight.10,13 The present study was underpowered to identify any association between low birth weight and children obesity. Small sample sizes and low prevalence of obesity in children with low birth weight may limit the statistical power in the analyses of associations between low birth weight and the risk of childhood obesity. These features directly address some of the primary reasons why earlier studies may have produced contrasting results. A recent Chinese study including very large samples (1703 children with low birth weight and 55 925 children in total) assessed the association of birth weight with the risk of obesity in children aged o 3 years.18 Results showed a positive association between birth weight and childhood obesity, indicating that low levels of birth weight, defined as birth weight o 2500 g, were associated with a decreased risk of childhood obesity, and higher © 2015 Macmillan Publishers Limited
Birth weight and childhood obesity Y Qiao et al
S77 Table 3.
Mean difference of BMI z-scores in 9–11–year-old children based on birth weight status P-value for trend
Birth weight (g) o 2500
2500–2999
3000–3499
3500–3999
⩾ 4000
Total Number of participants BMI z-score Multiple-adjusted BMI z-scoresb
350 − 0.05 (0.07) − 0.11 (0.08)
1050 0 0
1831 0.10 (0.05) 0.10 (0.05)
1421 0.26 (0.05) 0.25 (0.05)
489 0.44 (0.07) 0.47 (0.07)
o0.001 o0.001
Boys Number of participants BMI-for-age z-score Multiple-adjusted BMI z-scoreb
128 0.03 (0.13) − 0.04 (0.13)
444 0 0
777 0.05 (0.08) 0.06 (0.07)
708 0.17 (0.08) 0.15 (0.08)
282 0.39 (0.10) 0.43 (0.10)
o0.001 o0.001
Girls Number of participants BMI-for-age z-score Multiple-adjusted BMI z-scoreb
222 − 0.10 (0.09) −0.16 (0.10)
606 0 0
1054 0.14 (0.06) 0.14 (0.06)
713 0.32 (0.07) 0.34 (0.07)
207 0.47 (0.10) 0.50 (0.10)
o0.001 o0.001
a
Abbreviation: BMI, body mass index. Data are mean difference (s.e.m.). aAdjusted also for sex. bAdjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times.
Figure 1. OR of continuous birth weight in relation to its risk for obesity among boys (a), girls (b), and boys and girls combined (c). Adjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times. Lines with short dashes represent the pointwise 95% CIs. © 2015 Macmillan Publishers Limited
International Journal of Obesity Supplements (2015) S74 – S79
Birth weight and childhood obesity Y Qiao et al
S78 Table 4.
Odds ratio for obesity in 9–11-year-old children based on birth weight status P-value for trend
Birth weight (g) o 2500
2500–2999
3000–3499
3500–3999
⩾ 4000
Total Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
350 34 1.05 (0.69–1.59) 1.02 (0.65–1.59)
1050 101 1 1
1831 217 1.18 (0.91–1.52) 1.18 (0.91–1.54)
1421 201 1.45 (1.12–1.89) 1.45 (1.10–1.92)
489 86 1.91 (1.38–2.64) 2.08 (1.47–2.93)
o0.001 o0.001
Boys Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
128 21 1.38 (0.79–2.43) 1.37 (0.75–2.50)
444 55 1 1
777 114 1.08 (0.76–1.55) 1.08 (0.74–1.57)
708 118 1.34 (0.94–1.93) 1.28 (0.88–1.88)
282 52 1.59 (1.03–2.47) 1.77 (1.12–2.82)
0.154 0.093
Girls Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
222 13 0.74 (0.39–1.40) 0.71 (0.36–1.41)
606 46 1 1
1054 103 1.25 (0.86–1.80) 1.28 (0.88–1.88)
713 83 1.55 (1.06–2.29) 1.67 (1.11–2.50)
207 34 2.37 (1.45–3.86) 2.48 (1.47–4.17)
0.001 0.002
High-income countriesa Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
165 20 1.08 (0.61–1.91) 1.04 (0.56–1.94)
407 46 1 1
920 99 1.01 (0.69–1.47) 0.97 (0.65–1.43)
773 91 1.20 (0.81–1.76) 1.22 (0.81–1.83)
311 48 1.70 (1.09–2.66) 1.79 (1.11–2.89)
0.087 0.045
Low/middle-income countriesa Number of participants Number of cases Age-adjusted OR (95% CI) Multiple-adjusted OR (95% CI)b
185 14 0.94 (0.50–1.76) 0.91 (0.47–1.76)
643 55 1 1
911 118 1.32 (0.93–1.87) 1.33 (0.92–1.92)
648 110 1.68 (1.17–2.41) 1.61 (1.10–2.35)
178 38 2.07 (1.28–3.35) 2.30 (1.38–3.83)
o0.001 0.011
a
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio. aAdjusted also for sex. bAdjusted for highest parental education, maternal history of gestational diabetes, child age, infant feeding mode, gestational age, unhealthy diet pattern scores, healthy diet pattern scores, moderate-to-vigorous physical activity, sleeping and sedentary times.
levels of birth weight, defined as birth weight 43000 g, were associated with an increased risk of childhood obesity. To date, four studies have evaluated whether there is a sexspecific association between birth weight and the risk of childhood obesity.12,14,16,18 All four studies have found a positive association between birth weight and the risk of childhood obesity in girls. However, the associations between birth weight and the risk of childhood obesity in boys were less consistent. Three studies described a positive association between birth weight and the risk of childhood obesity,12,16,18 whereas one Chinese study found a U-shaped association between birth weight and the risk of childhood obesity.14 The research with Chinese children showed low levels of birth weight not to be associated with increased risk of childhood obesity in boys (OR 2.29; 95% CI: 0.80–6.50) or decreased risk of childhood obesity in girls (OR 0.66; 95% CI: 0.08–5.33).14 Such findings are consistent with those reported in the present study. Thus, more studies are needed to assess whether there is a sex difference in the association between birth weight and childhood obesity. The present study also indicated that the extent of high birth weight with the risk of childhood obesity was different in different countries. The relative risk of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and started to rise after 3500 g of birth weight in children from low/middle-income countries. A recent study found that higher levels of birth weight, defined as birth weight 43000 g, were associated with an increased risk of overweight or obesity among Chinese children from 6 months to 3 years of age (middle-income countries).18 Several studies in developed countries have shown higher levels of birth weight, defined as birth weight 44000 g, to International Journal of Obesity Supplements (2015) S74 – S79
be positively associated with an increased risk of childhood obesity.13,16 All these results seem to support our finding. The differences in maternal weight and nutritional status before and during pregnancy and child feeding between developed and developing countries might result in the different extent of association between high birth weight and the risk of childhood obesity between these countries. There are several strengths in our study including the recruitment of a large multinational sample of children from low- to high-income countries across several regions of the world, the highly standardized measurement protocol, the use of direct measurements whenever possible and the rigorous quality-control program. In addition, MVPA, sedentary behavior and sleep duration were objectively obtained from the 24-h, waist-worn accelerometer, and dietary intake was assessed by Food Frequency Questionnaire in the present study. There are also a number of limitations to the present study. First, the crosssectional study precludes us from making cause-and-effect inferences. Second, as the information on maternal prepregnancy BMI were not available in the present study, we may not be able to fully control for the effect of this variable on the association of birth weight with the risk of childhood obesity. Third, the birth weight records may have been recalled wrongly by the parent or guardian. The degree to which these factors may have biased the results is unknown. In summary, higher levels of birth weight, defined as birth weight ⩾ 3500 g, were positively associated with an increased risk of obesity in 9–11-year-old children in 12 countries. There was a U-shaped association between birth weight and the risk of childhood obesity among boys, and a positive linear association © 2015 Macmillan Publishers Limited
Birth weight and childhood obesity Y Qiao et al
S79 between birth weight and the risk of childhood obesity among girls. The relative risks of childhood obesity started to rise after 4000 g of birth weight in children from high-income countries, and after 3500 g of birth weight in children from low/middleincome countries. CONFLICT OF INTEREST MF has received a research grant from Fazer Finland and has received an honorarium for speaking for Merck. AK has been a member of the Advisory Boards of Dupont and McCain Foods. RK has received a research grant from Abbott Nutrition Research and Development. VM is a member of the Scientific Advisory Board of Actigraph and has received an honorarium for speaking for The Coca-Cola Company. TO has received an honorarium for speaking for The Coca-Cola Company. The remaining authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank the ISCOLE External Advisory Board and the ISCOLE participants and their families who made this study possible. A membership list of the ISCOLE Research Group and External Advisory Board is included in Katzmarzyk et al. (this issue). ISCOLE was funded by The Coca-Cola Company.
REFERENCES 1 World Health Organization. Obesity: preventing and managing the global epidemic. Geneva: World Health Organization; 2000. World Health Organ Tech Rep Ser 2013; 894: 1–252. 2 Wu JF. Childhood obesity: a growing global health hazard extending to adulthood. Pediatr Neonatol 2013; 54: 71–72. 3 Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2014; 384: 766–781. 4 Lu Y, Zhang J, Lu X, Xi W, Li Z. Secular trends of macrosomia in southeast China, 1994-2005. BMC Public Health 2011; 11: 818–818. 5 Koyanagi A, Zhang J, Dagvadorj A, Hirayama F, Shibuya K, Souza JP et al. Macrosomia in 23 developing countries: an analysis of a multicountry, facility-based, cross-sectional survey. Lancet 2013; 381: 476–483. 6 Schack-Nielsen L, Molgaard C, Sorensen TI, Greisen G, Michaelsen KF. Secular change in size at birth from 1973 to 2003: national data from Denmark. Obesity 2006; 14: 1257–1263. 7 Yu ZB, Han SP, Zhu GZ, Zhu C, Wang XJ, Cao XG et al. Birth weight and subsequent risk of obesity: a systematic review and meta-analysis. Obes Rev 2011; 12: 525–542. 8 Schellong K, Schulz S, Harder T, Plagemann A. 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. 9 Weng SF, Redsell SA, Swift JA, Yang M, Glazebrook CP. Systematic review and meta-analyses of risk factors for childhood overweight identifiable during infancy. Arch Dis Child 2012; 97: 1019–1026. 10 Frye C, Heinrich J. Trends and predictors of overweight and obesity in East German children. Int J Obes Relat Metab Disord 2003; 27: 963–969. 11 Monteiro PO, Victora CG, Barros FC, Monteiro LM. Birth size, early childhood growth, and adolescent obesity in a Brazilian birth cohort. Int J Obes Relat Metab Disord 2003; 27: 1274–1282. 12 Rugholm S, Baker JL, Olsen LW, Schack-Nielsen L, Bua J, Sorensen TI. Stability of the association between birth weight and childhood overweight during the development of the obesity epidemic. Obes Res 2005; 13: 2187–2194. 13 Dubois L, Girard M. Early determinants of overweight at 4.5 years in a populationbased longitudinal study. Int J Obes 2006; 30: 610–617.
© 2015 Macmillan Publishers Limited
14 Zhang X, Liu E, Tian Z, Wang W, Ye T, Liu G et al. High birth weight and overweight or obesity among Chinese children 3-6 years old. Prev Med 2009; 49: 172–178. 15 Ye R, Pei L, Ren A, Zhang Y, Zheng X, Liu J-m. Birth weight, maternal body mass index, and early childhood growth: a prospective birth cohort study in China. J Epidemiol 2010; 20: 421–428. 16 Oldroyd J, Renzaho A, Skouteris H. Low and high birth weight as risk factors for obesity among 4 to 5-year-old Australian children: does gender matter? Eur J Pediatr 2011; 170: 899–906. 17 Loaiza S, Coustasse A, Urrutia-Rojas X, Atalah E. Birth weight and obesity risk at first grade in a cohort of Chilean children. Nutr Hosp 2011; 26: 214–219. 18 Li N, Liu E, Sun S, Guo J, Pan L, Wang P et al. Birth weight and overweight or obesity risk in children under 3 years in China. Am J Hum Biol 2014; 26: 331–336. 19 Rooney BL, Mathiason MA, Schauberger CW. Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort. Matern Child Health J 2011; 15: 1166–1175. 20 Hart CN, Cairns A, Jelalian E. Sleep and obesity in children and adolescents. Pediatr Clin North Am 2011; 58: 715–733. 21 Pearson N, Biddle SJ. Sedentary behavior and dietary intake in children, adolescents, and adults. A systematic review. Am J Prev Med 2011; 41: 178–188. 22 Prentice-Dunn H, Prentice-Dunn S. Physical activity, sedentary behavior, and childhood obesity: a review of cross-sectional studies. Psychol Health Med 2012; 17: 255–273. 23 Katzmarzyk PT, Barreira TV, Broyles ST, Champagne CM, Chaput JP, Fogelholm M et al. The International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE): design and methods. BMC Public Health 2013; 13: 900. 24 Currie C, Gabhainn SN, Godeau E, Roberts C, Smith R, Currie D et al. (eds) Inequalities in Children's Health: HBSC International Report from the 2005/2006 Survey World Health Organization: Geneva, Switzerland, 2008. 25 Mikkilä V, Vepsäläinen H, Saloheimo T, Gonzalez SA, Meisel JD, Hu G et al. An international comparison of dietary patterns in 9-11-year old children. Int J Obes Suppl 2015 (this issue). 26 De Onis M, Onyanga AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull WHO 2007; 85: 660–667. 27 Tudor-Locke C, Mire EF, Dentro KN, Barreira TV, Schuna JM, Zhao P et al. A model for presenting accelerometer paradata in large studies: ISCOLE. Int J Behav Nutr Phys Act 2015; 12: 52. 28 Barreira TV, Schuna Jr JM, Mire EF, Katzmarzyk PT, Chaput JP, Leduc G et al. Identifying children's nocturnal sleep using 24-h waist accelerometry. Med Sci Sports Exerc 2015; 47: 937–943. 29 Tudor-Locke C, Barreira TV, Schuna JM, Mire EF, Chaput J-P, Fogelholm M et al. Improving wear time compliance with a 24-hour waist-worn accelerometer protocol in the International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE). Int J Behav Nutr Phys Act 2015; 12: 11. 30 Tudor-Locke C, Barreira TV, Schuna Jr JM, Mire EF, Katzmarzyk PT. Fully automated waist-worn accelerometer algorithm for detecting children's sleep-period time separate from 24-h physical activity or sedentary behaviors. Appl Physiol Nutr Me 2014; 39: 53–57. 31 Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sport Sci 2008; 26: 1557–1565. 32 Katzmarzyk PT, Barreira TV, Broyles ST, Champagne CM, Chaput J-P, Fogelholm M et al. Physical activity, sedentary time, and obesity in an international sample of children. Med Sci Sports Exerc 2015; 47: 2062–2069. 33 Eriksson JG. Early growth and adult health outcomes–lessons learned from the Helsinki Birth Cohort Study. Matern Child Nutr 2005; 1: 149–154. 34 Chen W, Srinivasan SR, Yao L, Li S, Dasmahapatra P, Fernandez C et al. Low birth weight is associated with higher blood pressure variability from childhood to young adulthood: the Bogalusa Heart Study. Am J Epidemiol 2012; 176 (Suppl 7): S99–S105.
International Journal of Obesity Supplements (2015) S74 – S79
