Stunting at 24 Months Is Not Related to Incidence of Overweight through Young Adulthood in an Urban South African Birth Cohort

Stunting at 24 Months Is Not Related to Incidence of Overweight through Young Adulthood in an... Abstract Background The role that childhood stunting plays in the development of overweight and obesity later in life is not well understood, particularly in adolescence and young adulthood, because most studies have only followed up through midchildhood. Objective The objective of this study was to examine the relation between stunting and age-specific patterns of overweight and obesity incidence from early childhood to young adulthood in the context of a country in the process of the nutrition transition while these children were growing up. Methods We analyzed data from 895 participants in the Birth-to-Twenty Plus Cohort (Bt20+), an urban South African birth cohort initiated in 1990. Anthropometric data were collected at multiple ages and participants were included if they provided height at age 24 mo and ≥1 measure of body mass index [BMI; weight (kg)/height (m)2] in each of the following time periods: 4–8 y, 11–12 y, 13–15 y, 16–18 y, and 22–24 y. We defined stunting at age 24 mo as height-for-age z score <2 and overweight as BMI z score (BMIZ) >1 in childhood (4–8 y) and adolescence (11–12 y, 13–15 y, and 16–18 y) and BMI ≥25 in young adulthood (22–24 y). We compared BMI, BMIZ, and the prevalence of overweight by stunting status, stratified by sex. Results Our sample was 93% black and 51% female. The prevalence of stunting at 24 mo was 26% in males and 19% in females. In young adulthood, the prevalence of overweight and obesity was 15.5% (men) and 47.5% (women). Among both males and females, neither mean BMI nor a combined measure of overweight and obesity in any subsequent period differed by stunting status at 24 mo (P ≥ 0.05). Conclusion Stunting at 24 mo was not related to the risk of overweight or obesity in this cohort. Stunting may not be an important contributor to the increasing obesity rates in urban South Africa. body mass index, overweight, stunting, children, South Africa Introduction Overweight and obesity are major global health concerns, because both increase the risk of chronic diseases such as cardiovascular disease and type 2 diabetes (1). The combined global prevalence of adult overweight and obesity has almost doubled since 1980, reaching nearly 38% in women and 37% in men (2). Childhood overweight and obesity are also rapidly becoming major global health problems. From 2009 to 2013 the combined prevalence of overweight and obesity in children ages 2–19 y increased by 47% globally (2). Childhood weight status has been shown to track into adulthood (3). Low- and middle-income countries (LMICs) are experiencing greater increases in both childhood and adult overweight and obesity rates relative to high-income countries (2, 4). At the same time, unlike in high-income countries, linear growth faltering due to chronic undernutrition in early life remains an important public health concern in many LMICs (5). As LMICs undergo transitions in economic, demographic, and nutritional conditions related to development and urbanization, they are faced with a dual burden of stunting and overweight (6, 7). This is particularly true in South Africa where stunting prevalence has remained relatively stagnant, whereas the prevalence of overweight and obesity rivals that of high-income countries (8–10). It has been suggested that childhood stunting is contributing to the high rates of adult obesity in LMICs (11). Nutritional insults in “the first 1000 days” from conception through the second birthday may permanently program an individual to increase or preserve fat stores, leading to the development of obesity if the food supply becomes abundant (12). Cross-sectional studies have reported a higher prevalence of overweight and obesity in stunted than in nonstunted children (11, 13). However, despite the plausible mechanisms and cross-sectional associations, there is a growing body of literature suggesting that, in some countries, when examined prospectively, there is a null or negative relation between childhood stunting and later obesity (14–20). Of these longitudinal studies, only 1, to our knowledge, has examined age-specific patterns of overweight and obesity incidence, and furthermore, only 1 has follow-up to adulthood (16, 19). Previous findings from the Birth-to-Twenty Plus Cohort (Bt20+), Africa's largest and longest running study of child and adolescent health and development, indicated no significant relation between stunting status at age 2 y and body composition at age 9 y (20). However, the relation between early childhood stunting and overweight and obesity in adolescence and adulthood in this cohort is still unexplored. This study aimed to extend the follow-up period and examine the relation between early-childhood stunting and age-specific patterns of overweight and obesity from early childhood to young adulthood. The objectives of this study are to examine differences in the prevalence and incidence of overweight and obesity at selected ages by stunting status at 24 mo in an urban South African birth cohort. Methods We analyzed data from participants in the Bt20+, a longitudinal study of singleton children born between April and June 1990 in Soweto and Johannesburg, South Africa. The children were followed up 21 times from birth to age 19 y, and subsequently into young adulthood (age 22–24 y). Detailed information on the cohort is provided elsewhere (21). Of all births notified in the study area, 3273 children who were born in and who remained residents of the Johannesburg-Soweto area for 6 mo were recruited. The cohort has experienced a relatively low attrition rate of 30% over 20 y, with most of the attrition occurring in the first few years of the study, primarily due to migration out of the study area (21). The sample is representative of urban South Africa, with 78% of participants being black African and approximately equal numbers of males (49%) and females (51%), predominantly of low socioeconomic status. Participants or their caregiver gave written informed consent, and ethical approval was obtained from the University of the Witwatersrand Committee for Research on Human Subjects (approval ID M010556). Anthropometric measurements including height and weight were collected with the use of standard protocols at all time points. Child characteristics at birth (birth weight and gestational age) were collected from birth notification forms or hospital records. Maternal characteristics (height, education, socioeconomic status) were collected by interview between birth and child age of 24 mo. To assess socioeconomic status, an asset score ranging from 0 to 7 was calculated with the use of variables for home type, home ownership, electricity in the home, and ownership of a television, car, refrigerator, washing machine, and phone; this score was then grouped into quintiles. Height-for-age z score (HAZ) was calculated by using the WHO child growth reference (22, 23). Stunting was defined as HAZ <2. BMI was defined as weight (kilograms) divided by height (meters) squared (kg/m2). BMI values were converted to z scores (BMIZ) by using the WHO reference (22, 23). We defined combined overweight and obesity with the use of the WHO reference for children aged 5–19 y and the WHO cutoffs for adults (23–25). For the childhood and adolescent time periods, combined overweight and obesity was defined as BMIZ >1.0 SD (24, 26). Although the WHO defines overweight and obesity for children aged 0–4 y as BMIZ >2.0 SDs, for consistency we used the WHO definition for children aged 5–19 y for all childhood and adolescent time periods (22–24). For the young adult time period, combined overweight and obesity was defined as BMI ≥ 25 (25, 27). Not all of the participants had data collected each year. For this reason, as described elsewhere (28), we grouped survey rounds into the following time periods to maximize available data: age 2 y, early childhood (4–8 y), early adolescence (11–12 y), mid-adolescence (13–15 y), late adolescence (16–18 y), and young adulthood (22–24 y). When a participant had >1 BMI value recorded during a time period, 1 value was chosen at random to represent that time period. To better represent the population of interest, we restricted analyses to participants who were black, Indian, or of mixed ancestral heritage. For the final analytic sample, we included participants who had a length measurement recorded at age 24 mo and ≥1 BMI value in each of the 6 time periods (n = 895). A flow chart depicting the final sample of eligible children included in the analyses is provided in Supplemental Figure 1. We calculated prevalence in each time period as the percentage of participants who were overweight or obese in that time period. We calculated incidence of overweight or obesity as the percentage of participants who were not overweight or obese in the previous time period but became overweight or obese in the subsequent time period. We calculated incidence density rate as the number of incident cases per 1000 person-years, with person years being the sum of the number of years of follow-up contributed by each participant who was at risk of becoming overweight or obese. We conducted all of the analyses separately for males and females, due to previously shown sex differences in obesity incidence in this population (28). In addition, to ensure adequately sized strata, we examined the prevalence and incidence of overweight and obesity combined. We tested differences in selected child, maternal, and household characteristics by sex and stunting status at 24 mo with the use of Student's t tests and Pearson's chi-square tests. We compared sex-stratified mean BMI and BMIZ and the prevalence of combined overweight and obesity in each time period by stunting status at 24 mo with the use of Student's t tests and Pearson's chi-square tests, respectively. We compared sex-stratified combined overweight and obesity incidence density rates by stunting status at 24 mo in each time period by calculating incidence density ratios and 95% CIs. Differences in means, frequencies, and rates were considered significant at P < 0.05. We conducted a sensitivity analysis to examine whether cohort participants in the final analytic sample (i.e., those who had length measured at age 2 y and BMI data for all 6 time periods) were significantly different from the individuals who had a length measurement at age 2 y but who were excluded from our analysis because they did not provide BMI data for all 6 time periods. We compared the prevalence of stunting at 24 mo and the prevalence of combined overweight and obesity in each time period by inclusion status with the use of Pearson's chi-square tests. Differences in prevalence were considered significant at P < 0.05. We used SAS version 9.4 (SAS Institute, Inc.) for all analyses. Results The large majority of participants were black (93%). Nearly 26% of males (112 of 438) and 19% of females (88 of 457) in this sample were stunted at 24 mo of age. HAZ and growth patterns over time for this population have been reported elsewhere (29). Birth weight, birth weight z score, HAZ at 24 mo, and maternal height were lower (all P < 0.05) in those who were stunted at 24 mo than in those who were not stunted at 24 mo in both males and females (Table 1). BMI and BMIZ at 24 mo were significantly higher among those who were stunted at 24 mo compared with those who were not stunted at 24 mo in both males and females. The percentage of participants in the lowest-asset quintile did not significantly differ by stunting status at 24 mo in males or females. TABLE 1 Characteristics of South African children by sex and stunting status at 24 mo: Bt20+ cohort1 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 1Values are means ± SDs unless otherwise indicated, n = 895. “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score. View Large TABLE 1 Characteristics of South African children by sex and stunting status at 24 mo: Bt20+ cohort1 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 1Values are means ± SDs unless otherwise indicated, n = 895. “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score. View Large Among both males and females, the mean BMI in each of the subsequent time periods did not significantly differ by stunting status at 24 mo (Table 2). Mean BMIZs in early adolescence (11–12 y) and in mid-adolescence (13–15 y) were significantly lower in males who were stunted at 24 mo compared with males who were not stunted at 24 mo (both P < 0.05). Mean BMIZ in early adolescence (11–12 y) was significantly lower among females who were stunted at 24 mo compared with females who were not stunted at 24 mo (P = 0.03). The mean BMIZ at each measurement point by sex and stunting status at 24 mo is shown in Supplemental Figure 2, and the mean BMI at each measurement point by sex and stunting status at 24 mo is shown in Supplemental Figure 3. TABLE 2 BMI, BMIZ, and prevalence of overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 1Values are means ± SDs unless otherwise indicated. ``Overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 2-, 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. View Large TABLE 2 BMI, BMIZ, and prevalence of overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 1Values are means ± SDs unless otherwise indicated. ``Overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 2-, 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. View Large In young adulthood (22–24 y), the prevalence of overweight and obesity was 15.5% in males and 47.5% in females. The increased incidence of overweight and obesity in the females of this cohort has been reported previously (28). Although not widely prevalent among males in any time period, combined overweight and obesity increased over time in females, reaching nearly 41% by young adulthood among those who were stunted at 24 mo and 49% among those who were not stunted at 24 mo. The prevalence of combined overweight and obesity in any of the time periods did not significantly differ by stunting status at 24 mo among either males or females (Table 2). In male participants, the period incidence and incidence density rate of combined overweight and obesity were highest between late adolescence (16–18 y) and young adulthood (22–24 y) (Table 3). These rates did not differ significantly in any time period by stunting status at 24 mo. No males who were stunted at 24 mo became overweight or obese between early adolescence (11–12 y) and mid-adolescence (13–15 y). In females, the period incidence of combined overweight and obesity was highest between late adolescence (16–18 y) and young adulthood (22–24 y), but the incidence density rate was highest between early adolescence (11–12 y) and mid-adolescence (13–15 y). These rates did not differ in any time period by stunting status at 24 mo (Table 3). TABLE 3 Incidence of combined overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 1``Combined overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. 2Incident overweight or obesity cases during the period among those who were at risk of overweight or obesity at the beginning of the period. 3Number of incident overweight or obesity cases during the period per 1000 person-years. 4Rate ratio comparing the incidence of combined overweight and obesity in those who were stunted at 24 mo with those who were not stunted at 24 mo. View Large TABLE 3 Incidence of combined overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 1``Combined overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. 2Incident overweight or obesity cases during the period among those who were at risk of overweight or obesity at the beginning of the period. 3Number of incident overweight or obesity cases during the period per 1000 person-years. 4Rate ratio comparing the incidence of combined overweight and obesity in those who were stunted at 24 mo with those who were not stunted at 24 mo. View Large Among participants with a length measurement at age 2 y, the prevalence of stunting at 24 mo in those included in the analyses was not significantly different from that of those who were not included, in either males or females. In males with a length measurement at age 2 y, the prevalence of combined overweight and obesity in mid-adolescence (13–15 y) was significantly different between these 2 groups (P = 0.03) but not in any other time period (Supplemental Table 1). In females with a length measurement at age 2 y, there were no significant differences in the prevalence of combined overweight and obesity between those included in the analyses and those not included in any time period. The distribution of asset quintile significantly differed between participants included in the analyses and participants not included in the analyses (P < 0.001 in both males and females). Between the late-adolescent (16–18 y) and young adult (22–24 y) time periods, 277 participants were lost to follow-up. The prevalence of overweight and obesity did not significantly differ in any earlier time period between participants included in the analyses and participants lost to follow-up after late adolescence. Discussion In this urban South African birth cohort of males and females, we examined the prevalence and incidence of combined overweight and obesity during 5 time periods from early childhood to young adulthood and found no significant differences by stunting status at 24 mo. In LMICs, prospective data are scarce. A strength of our study was the use of longitudinal data spanning the period from early childhood to adulthood. Given the increasing rates of overweight and obesity and related morbidities in LMICs, it is important to understand factors contributing to this trend. Our findings suggest that stunting at 24 mo may not be a significant factor contributing to the increasing overweight and obesity rates in urban South Africa, and therefore that there must be other mechanisms that warrant further research. Previous research has suggested that stunting in childhood may increase the risk of obesity later in life. This association has been shown previously in Russia, China, and Brazil, as well as in South Africa, which are all countries in various stages of nutrition transition (11, 13). An underlying mechanism for this relation is thought to be long-term impaired fat oxidation, a known risk factor for excess weight gain (30, 31). Most studies on this relation have been cross-sectional, and therefore causality has not been established. The relation between childhood stunting and overweight and obesity in South Africa has been examined previously, with mixed findings. Our results at 24 mo affirm previous cross-sectional findings from other transitioning countries, that, at 24 mo, children who were stunted had higher BMI, BMIZ, and prevalence of overweight and obesity than children who were not stunted. However, this association may be spurious and due to random errors in height measurement. Steyn et al. (32), with the use of data from the South African National Food Consumption Survey (NFCS) conducted in 1999, reported that children aged 1–9 y who were short for their age had a higher risk of being overweight. Similarly, Symington et al. (33) with the use of data from the second South African NFCS–Fortification Baseline I conducted in 2005, reported a significant inverse association between BMI and HAZ in South African children aged 3–10 y. On the contrary, Jinabhai et al. (34), using primary data from a community survey in rural South Africa and secondary provincial data from the South African National Primary Schools' Anthropometric (SANPS) Survey in 1994–1995, found no relation between stunting and overweight in children aged 8–11 y, and Mukuddem-Petersen and Kruger (35) found no association between stunting and overweight in schoolchildren aged 10–15 y of age of the North West Province in 2005. A growing body of literature suggests that, when examined longitudinally, there is a null or negative relation between childhood stunting and later overweight and obesity. We extended earlier findings in this same cohort by Cameron and Getz (36), who found no significant relation between stunting status at 24 mo and detailed body composition at 9 y. Outcomes for that study were only collected in late childhood, and the authors suggested that the lack of a relation could be due to the cohort not yet having reached puberty. Our study addressed this gap by using BMI data spanning early childhood to young adulthood. Our findings are consistent with those of Cameron et al. (20) in the childhood time periods. Furthermore, we found no evidence that those who were stunted at 24 mo were any more likely to become overweight or obese in the adolescent and young-adult time periods than those who were not stunted at 24 mo. Our study, however, only used BMI, rather than detailed body-composition data. The findings of our study add to the evidence of a null or negative association between childhood stunting and future overweight and obesity (14–20). Of the longitudinal studies, only 1 study has looked at age-specific patterns of overweight and obesity and found stunting at 6–18 mo to be associated with reduced prevalence of overweight later in childhood in a cohort of Peruvian children (19). The inverse association in that study compared with the null findings in ours could be due to the different ages (6–18 mo compared with 24 mo) used to characterize children as stunted. Like the Peruvian study, we looked at outcomes at numerous follow-up ages; however, a strength of our study was the inclusion of a follow-up point in adulthood. To our knowledge, only 1 other study has prospectively looked at the relation between childhood stunting and adult overweight and obesity (16). Schroeder et al. (16) found an association between childhood stunting and lower BMI and percentage body fat in men, but no association in women in Guatemala. However, after controlling for BMI and percentage body fat, Schroeder et al. found a positive association between stunting and increased abdominal fat in both sexes. The use of BMI in our study to classify overweight and obesity has limitations. BMI is a measure of weight in proportion to height and does not distinguish between lean mass and fat mass. In addition, in using only BMI we are unable to comment on the distribution of subcutaneous fat, which is known to be highly related to morbidity (37). There is evidence to suggest that the trajectory of muscle growth is established prenatally, and that rapid weight gain after undernutrition may lead children to have proportionately less lean muscle tissue and more fat tissue than a child with similar age and weight-for-height who was not undernourished (38, 39). Although we found no association between stunting at age 2 y and BMI in any time period in our study, in a separate subsample of Bt20+ participants, Kagura et al. (40) found an association between stunting at age 2 y and lower lean mass at age 10 y. Future research should aim to examine the relation between stunting in early life and detailed body composition and fat distribution in adulthood. In 2 time periods for males and 1 time period for females, BMIZ significantly differed between those who were stunted at 24 mo and those who were not stunted at 24 mo. However, all differences in BMIZ were in the same direction, and at no point was the BMIZ of those who were stunted at 24 mo higher than the BMIZ of those who were not stunted at 24 mo in either males or females. Differences seen in BMIZ but not BMI were likely due to the process of standardization. By standardizing to a z score we reduced the variability in the measure of BMI. Lower variability increases the ability to detect statistical significance and may have accentuated the differences in BMIZ by stunting status. Our analytical sample was restricted to the 895 participants who had a length measurement at 2 y and a BMI value for each of the 6 time periods. Stratifying the analytic sample by both sex and stunting status resulted in small group sizes, particularly in males, of whom few became overweight or obese in childhood and adolescence. Our analyses may have been underpowered to detect small differences in incidence density within specific age and sex strata. We found no meaningful significant differences in the prevalence of stunting at 24 mo or the prevalence of overweight or obesity in any time period when those with complete longitudinal data were compared with those with incomplete data, leading us to believe that selection bias does not explain our findings. However, our analytic sample had a lower percentage of participants in the poorest-asset quintile compared with those not included in the analyses, so caution must be used when interpreting results. For consistency, we applied the WHO cutoffs for children aged 5–19 y for all childhood and adolescent time periods, which would have overestimated the prevalence of overweight and obesity at age 2 y and, to a lesser extent, at age 4–8 y. However, this would not have affected comparisons between stunted and nonstunted children within any period. Our study adds to the limited evidence on whether childhood stunting is contributing to the increasing rates of overweight and obesity in LMICs and addresses a gap in the literature with regard to follow-up in adolescence and adulthood. We found no evidence that the prevalence or incidence of overweight or obesity in any time period from early childhood to young adulthood differed by stunting at 24 mo in this urban South African cohort. This suggests that stunting may not be an important contributor to the increasing childhood or adult obesity rates in urban South Africa. Acknowledgments The authors’ responsibilities were as follows—SKH, EAL, and ADS: designed the study; SKH: conducted the data analysis with contributions from RJM and EAL; SKH: wrote the first draft of manuscript; LMR and SAN: were the principal investigators of the Birth-to-Twenty Study in South Africa and provided access to the data; RJM, SAN, and ADS: contributed to interpreting the results and writing the final manuscript; and all authors: read and approved the final manuscript. Notes Supported by NIH research training grant R25 TW009337, funded by the Fogarty International Center and the National Institute of Mental Health, and the Department of Science and Technology and National Research Foundation Centre of Excellence in Human Development at the University of the Witwatersrand. The Birth-to-Twenty Study received support from the Wellcome Trust (United Kingdom), the South African Medical Research Council, the University of the Witwatersrand, Johannesburg, the Human Sciences Research Council, and the United Kingdom Medical Research Council/ Department for International Development Africa Research Leader Scheme. Author disclosures: SKH, RJM, EAL, LLM, SAN, and ADS, no conflicts of interest. Supplemental Figures 1–3 and Supplemental Table 1 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn/. Abbreviations used: BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; LMIC, low- and middle-income country. References 1. Guh DP , Zhang W , Bansback N , Amarsi Z , Birmingham CL , Anis AH . The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis . BMC Public Health 2009 ; 9 : 88 . Google Scholar CrossRef Search ADS PubMed 2. Ng M , Fleming T , Robinson M , Thomson B , Graetz N , Margono C , Mullany EC , Biryukov S , Abbafati C , Abera SF 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 – 81 . Google Scholar CrossRef Search ADS PubMed 3. Dietz WH . Critical periods in childhood for the development of obesity . Am J Clin Nutr 1994 ; 59 : 955 – 9 . Google Scholar CrossRef Search ADS PubMed 4. de Onis M , Blossner M , Borghi E . Global prevalence and trends of overweight and obesity among preschool children . Am J Clin Nutr 2010 ; 92 : 1257 – 64 . Google Scholar CrossRef Search ADS PubMed 5. de Onis M , Blossner M , Borghi E . Prevalence and trends of stunting among pre-school children, 1990–2020 . Public Health Nutr 2012 ; 15 : 142 – 8 . Google Scholar CrossRef Search ADS PubMed 6. Tzioumis E , Adair LS . Childhood dual burden of under- and overnutrition in low- and middle-income countries: a critical review . Food Nutr Bull 2014 ; 35 : 230 – 43 . Google Scholar CrossRef Search ADS PubMed 7. Popkin BM . The nutrition transition and its health implications in lower-income countries . Public Health Nutr 1998 ; 1 : 5 – 21 . Google Scholar CrossRef Search ADS PubMed 8. Said-Mohamed R , Micklesfield LK , Pettifor JM , Norris SA . Has the prevalence of stunting in South African children changed in 40 years? A systematic review . BMC Public Health 2015 ; 15 : 534 . Google Scholar CrossRef Search ADS PubMed 9. Pettifor JM . Combined stunting and overweight in young children—a paradox? South Afr J Clin Nutr 2006 ; 19 : 98 – 100 . Google Scholar CrossRef Search ADS PubMed 10. Bourne LT , Lambert EV , Steyn K . Where does the black population of South Africa stand on the nutrition transition? Public Health Nutr 2002 ; 5 : 157 – 62 . Google Scholar CrossRef Search ADS PubMed 11. Popkin BM , Richards MK , Montiero CA . Stunting is associated with overweight in children of four nations that are undergoing the nutrition transition . J Nutr 1996 ; 126 : 3009 – 16 . Google Scholar CrossRef Search ADS PubMed 12. Barker DJ . The fetal and infant origins of adult disease . BMJ 1990 ; 301 : 1111 . Google Scholar CrossRef Search ADS PubMed 13. Sawaya AL , Dallal G , Solymos G , de Sousa MH , Ventura ML , Roberts SB , Sigulem DM . Obesity and malnutrition in a Shantytown population in the city of Sao Paulo, Brazil . Obes Res 1995 ; 3 ( Suppl 2 ): 107s – 15s . Google Scholar CrossRef Search ADS PubMed 14. Walker SP , Gaskin PS , Powell CA , Bennett FI . The effects of birth weight and postnatal linear growth retardation on body mass index, fatness and fat distribution in mid and late childhood . Public Health Nutr 2002 ; 5 : 391 – 6 . Google Scholar CrossRef Search ADS PubMed 15. Walker SP , Chang SM , Powell CA . The association between early childhood stunting and weight status in late adolescence . Int J Obes (Lond) 2007 ; 31 : 347 – 52 . Google Scholar CrossRef Search ADS PubMed 16. Schroeder DG , Martorell R , Flores R . Infant and child growth and fatness and fat distribution in Guatemalan adults . Am J Epidemiol 1999 ; 149 : 177 – 85 . Google Scholar CrossRef Search ADS PubMed 17. Benefice E , Garnier D , Simondon KB , Malina RM . Relationship between stunting in infancy and growth and fat distribution during adolescence in Senegalese girls . Eur J Clin Nutr 2001 ; 55 : 50 – 58 . Google Scholar CrossRef Search ADS PubMed 18. Tanner S , Leonard WR , Reyes-Garcia V . The consequences of linear growth stunting: influence on body composition among youth in the Bolivian Amazon . Am J Phys Anthropol 2014 ; 153 : 92 – 102 . Google Scholar CrossRef Search ADS PubMed 19. Andersen CT , Stein AD , Reynolds SA , Behrman JR , Crookston BT , Dearden KA , Penny ME , Schott W , Fernald LCH . Stunting in infancy is associated with decreased risk of high body mass index for age at 8 and 12 years of age . J Nutr 2016 ; 146 : 2296 – 303 . Google Scholar CrossRef Search ADS PubMed 20. Cameron N , Wright MM , Griffiths PL , Norris SA , Pettifor JM . Stunting at 2 years in relation to body composition at 9 years in African urban children . Obes Res 2005 ; 13 : 131 – 6 . Google Scholar CrossRef Search ADS PubMed 21. Richter L , Norris S , Pettifor J , Yach D , Cameron N . Cohort profile: Mandela's children: the 1990 Birth to Twenty Study in South Africa . Int J Epidemiol 2007 ; 36 : 504 – 11 . Google Scholar CrossRef Search ADS PubMed 22. WHO Multicentre Growth Reference Study Group . WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development . Geneva (Switzerland) : WHO ; 2006 . 23. de Onis M , Onyango AW , Borghi E , Siyam A , Nishida C , Siekmann J . Development of a WHO growth reference for school-aged children and adolescents . Bull World Health Organ 2007 ; 85 : 660 – 7 . Google Scholar CrossRef Search ADS PubMed 24. Cole TJ , Bellizzi MC , Flegal KM , Dietz WH . Establishing a standard definition for child overweight and obesity worldwide: international survey . BMJ 2000 ; 320 : 1240 – 3 . Google Scholar CrossRef Search ADS PubMed 25. WHO . Obesity: preventing and managing the global epidemic . Report of a WHO consultation . World Health Organ Tech Rep Ser 2000 ; 894 : i – xii , 1–253 . PubMed 26. Van den Broeck J , Willie D , Younger N . The World Health Organization child growth standards: expected implications for clinical and epidemiological research . Eur J Pediatr 2009 ; 168 : 247 – 51 . Google Scholar CrossRef Search ADS PubMed 27. de Onis M , Habicht JP . Anthropometric reference data for international use: recommendations from a World Health Organization Expert Committee . Am J Clin Nutr 1996 ; 64 : 650 – 8 . Google Scholar CrossRef Search ADS PubMed 28. Lundeen EA , Norris SA , Adair LS , Richter LM , Stein AD . Sex differences in obesity incidence: 20-year prospective cohort in South Africa . Pediatr Obes 2016 ; 11 ( 1 ): 75 – 80 . Google Scholar CrossRef Search ADS PubMed 29. Stein AD , Wang M , Martorell R , Norris SA , Adair LS , Bas I , Sachdev HS , Bhargava SK , Fall CH , Gigante DP et al. Growth patterns in early childhood and final attained stature: data from five birth cohorts from low- and middle-income countries . Am J Hum Biol 2010 ; 22 : 353 – 9 . Google Scholar CrossRef Search ADS PubMed 30. Hoffman DJ , Sawaya AL , Verreschi I , Tucker KL , Roberts SB . Why are nutritionally stunted children at increased risk of obesity? Studies of metabolic rate and fat oxidation in shantytown children from Sao Paulo, Brazil . Am J Clin Nutr 2000 ; 72 : 702 – 7 . Google Scholar CrossRef Search ADS PubMed 31. Sawaya AL , Grillo LP , Verreschi I , da Silva AC , Roberts SB . Mild stunting is associated with higher susceptibility to the effects of high fat diets: studies in a shantytown population in Sao Paulo, Brazil . J Nutr 1998 ; 128 ( Suppl ): 415s – 20s . Google Scholar CrossRef Search ADS PubMed 32. Steyn NP , Labadarios D , Maunder E , Nel J , Lombard C . Secondary anthropometric data analysis of the National Food Consumption Survey in South Africa: the double burden . Nutrition 2005 ; 21 : 4 – 13 . Google Scholar CrossRef Search ADS PubMed 33. Symington EA , Gericke GJ , Nel JH , Labadarios D . The relationship between stunting and overweight among children from South Africa: secondary analysis of the National Food Consumption Survey–Fortification Baseline I . S Afr Med J 2015 ; 106 : 65 – 69 . Google Scholar CrossRef Search ADS PubMed 34. Jinabhai CC , Taylor M , Sullivan KR . Implications of the prevalence of stunting, overweight and obesity amongst South African primary school children: a possible nutritional transition? Eur J Clin Nutr 2003 ; 57 : 358 – 65 . Google Scholar CrossRef Search ADS PubMed 35. Mukuddem-Petersen J , Kruger HS . Association between stunting and overweight among 10-15-y-old children in the North West Province of South Africa: the THUSA BANA study . Int J Obes Relat Metab Disord 2004 ; 28 : 842 – 51 . Google Scholar CrossRef Search ADS PubMed 36. Cameron N , Getz B . Sex differences in the prevalence of obesity in rural African adolescents . Int J Obes Relat Metab Disord 1997 ; 21 : 775 – 82 . Google Scholar CrossRef Search ADS PubMed 37. Botton J , Heude B , Kettaneh A , Borys JM , Lommez A , Bresson JL , Ducimetiere P , Charles MA . Cardiovascular risk factor levels and their relationships with overweight and fat distribution in children: the Fleurbaix Laventie Ville Sante II Study . Metabolism 2007 ; 56 : 614 – 22 . Google Scholar CrossRef Search ADS PubMed 38. Singhal A , Wells J , Cole TJ , Fewtrell M , Lucas A . Programming of lean body mass: a link between birth weight, obesity, and cardiovascular disease? Am J Clin Nutr 2003 ; 77 : 726 – 30 . Google Scholar CrossRef Search ADS PubMed 39. Sawaya AL , Roberts S . Stunting and future risk of obesity: principal physiological mechanisms . Cad Saude Publica 2003 ; 19 ( Suppl 1 ): S21 – 8 . Google Scholar CrossRef Search ADS PubMed 40. Kagura J , Feeley AB , Micklesfield LK , Pettifor JM , Norris SA . Association between infant nutrition and anthropometry, and pre-pubertal body composition in urban South African children . J Dev Orig Health Dis 2012 ; 3 : 415 – 23 . Google Scholar CrossRef Search ADS PubMed © 2018 American Society for Nutrition. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Nutrition Oxford University Press

Stunting at 24 Months Is Not Related to Incidence of Overweight through Young Adulthood in an Urban South African Birth Cohort

Loading next page...
 
/lp/ou_press/stunting-at-24-months-is-not-related-to-incidence-of-overweight-nmJ9ZuV0gp
Publisher
Oxford University Press
Copyright
© 2018 American Society for Nutrition.
ISSN
0022-3166
eISSN
1541-6100
D.O.I.
10.1093/jn/nxy061
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The role that childhood stunting plays in the development of overweight and obesity later in life is not well understood, particularly in adolescence and young adulthood, because most studies have only followed up through midchildhood. Objective The objective of this study was to examine the relation between stunting and age-specific patterns of overweight and obesity incidence from early childhood to young adulthood in the context of a country in the process of the nutrition transition while these children were growing up. Methods We analyzed data from 895 participants in the Birth-to-Twenty Plus Cohort (Bt20+), an urban South African birth cohort initiated in 1990. Anthropometric data were collected at multiple ages and participants were included if they provided height at age 24 mo and ≥1 measure of body mass index [BMI; weight (kg)/height (m)2] in each of the following time periods: 4–8 y, 11–12 y, 13–15 y, 16–18 y, and 22–24 y. We defined stunting at age 24 mo as height-for-age z score <2 and overweight as BMI z score (BMIZ) >1 in childhood (4–8 y) and adolescence (11–12 y, 13–15 y, and 16–18 y) and BMI ≥25 in young adulthood (22–24 y). We compared BMI, BMIZ, and the prevalence of overweight by stunting status, stratified by sex. Results Our sample was 93% black and 51% female. The prevalence of stunting at 24 mo was 26% in males and 19% in females. In young adulthood, the prevalence of overweight and obesity was 15.5% (men) and 47.5% (women). Among both males and females, neither mean BMI nor a combined measure of overweight and obesity in any subsequent period differed by stunting status at 24 mo (P ≥ 0.05). Conclusion Stunting at 24 mo was not related to the risk of overweight or obesity in this cohort. Stunting may not be an important contributor to the increasing obesity rates in urban South Africa. body mass index, overweight, stunting, children, South Africa Introduction Overweight and obesity are major global health concerns, because both increase the risk of chronic diseases such as cardiovascular disease and type 2 diabetes (1). The combined global prevalence of adult overweight and obesity has almost doubled since 1980, reaching nearly 38% in women and 37% in men (2). Childhood overweight and obesity are also rapidly becoming major global health problems. From 2009 to 2013 the combined prevalence of overweight and obesity in children ages 2–19 y increased by 47% globally (2). Childhood weight status has been shown to track into adulthood (3). Low- and middle-income countries (LMICs) are experiencing greater increases in both childhood and adult overweight and obesity rates relative to high-income countries (2, 4). At the same time, unlike in high-income countries, linear growth faltering due to chronic undernutrition in early life remains an important public health concern in many LMICs (5). As LMICs undergo transitions in economic, demographic, and nutritional conditions related to development and urbanization, they are faced with a dual burden of stunting and overweight (6, 7). This is particularly true in South Africa where stunting prevalence has remained relatively stagnant, whereas the prevalence of overweight and obesity rivals that of high-income countries (8–10). It has been suggested that childhood stunting is contributing to the high rates of adult obesity in LMICs (11). Nutritional insults in “the first 1000 days” from conception through the second birthday may permanently program an individual to increase or preserve fat stores, leading to the development of obesity if the food supply becomes abundant (12). Cross-sectional studies have reported a higher prevalence of overweight and obesity in stunted than in nonstunted children (11, 13). However, despite the plausible mechanisms and cross-sectional associations, there is a growing body of literature suggesting that, in some countries, when examined prospectively, there is a null or negative relation between childhood stunting and later obesity (14–20). Of these longitudinal studies, only 1, to our knowledge, has examined age-specific patterns of overweight and obesity incidence, and furthermore, only 1 has follow-up to adulthood (16, 19). Previous findings from the Birth-to-Twenty Plus Cohort (Bt20+), Africa's largest and longest running study of child and adolescent health and development, indicated no significant relation between stunting status at age 2 y and body composition at age 9 y (20). However, the relation between early childhood stunting and overweight and obesity in adolescence and adulthood in this cohort is still unexplored. This study aimed to extend the follow-up period and examine the relation between early-childhood stunting and age-specific patterns of overweight and obesity from early childhood to young adulthood. The objectives of this study are to examine differences in the prevalence and incidence of overweight and obesity at selected ages by stunting status at 24 mo in an urban South African birth cohort. Methods We analyzed data from participants in the Bt20+, a longitudinal study of singleton children born between April and June 1990 in Soweto and Johannesburg, South Africa. The children were followed up 21 times from birth to age 19 y, and subsequently into young adulthood (age 22–24 y). Detailed information on the cohort is provided elsewhere (21). Of all births notified in the study area, 3273 children who were born in and who remained residents of the Johannesburg-Soweto area for 6 mo were recruited. The cohort has experienced a relatively low attrition rate of 30% over 20 y, with most of the attrition occurring in the first few years of the study, primarily due to migration out of the study area (21). The sample is representative of urban South Africa, with 78% of participants being black African and approximately equal numbers of males (49%) and females (51%), predominantly of low socioeconomic status. Participants or their caregiver gave written informed consent, and ethical approval was obtained from the University of the Witwatersrand Committee for Research on Human Subjects (approval ID M010556). Anthropometric measurements including height and weight were collected with the use of standard protocols at all time points. Child characteristics at birth (birth weight and gestational age) were collected from birth notification forms or hospital records. Maternal characteristics (height, education, socioeconomic status) were collected by interview between birth and child age of 24 mo. To assess socioeconomic status, an asset score ranging from 0 to 7 was calculated with the use of variables for home type, home ownership, electricity in the home, and ownership of a television, car, refrigerator, washing machine, and phone; this score was then grouped into quintiles. Height-for-age z score (HAZ) was calculated by using the WHO child growth reference (22, 23). Stunting was defined as HAZ <2. BMI was defined as weight (kilograms) divided by height (meters) squared (kg/m2). BMI values were converted to z scores (BMIZ) by using the WHO reference (22, 23). We defined combined overweight and obesity with the use of the WHO reference for children aged 5–19 y and the WHO cutoffs for adults (23–25). For the childhood and adolescent time periods, combined overweight and obesity was defined as BMIZ >1.0 SD (24, 26). Although the WHO defines overweight and obesity for children aged 0–4 y as BMIZ >2.0 SDs, for consistency we used the WHO definition for children aged 5–19 y for all childhood and adolescent time periods (22–24). For the young adult time period, combined overweight and obesity was defined as BMI ≥ 25 (25, 27). Not all of the participants had data collected each year. For this reason, as described elsewhere (28), we grouped survey rounds into the following time periods to maximize available data: age 2 y, early childhood (4–8 y), early adolescence (11–12 y), mid-adolescence (13–15 y), late adolescence (16–18 y), and young adulthood (22–24 y). When a participant had >1 BMI value recorded during a time period, 1 value was chosen at random to represent that time period. To better represent the population of interest, we restricted analyses to participants who were black, Indian, or of mixed ancestral heritage. For the final analytic sample, we included participants who had a length measurement recorded at age 24 mo and ≥1 BMI value in each of the 6 time periods (n = 895). A flow chart depicting the final sample of eligible children included in the analyses is provided in Supplemental Figure 1. We calculated prevalence in each time period as the percentage of participants who were overweight or obese in that time period. We calculated incidence of overweight or obesity as the percentage of participants who were not overweight or obese in the previous time period but became overweight or obese in the subsequent time period. We calculated incidence density rate as the number of incident cases per 1000 person-years, with person years being the sum of the number of years of follow-up contributed by each participant who was at risk of becoming overweight or obese. We conducted all of the analyses separately for males and females, due to previously shown sex differences in obesity incidence in this population (28). In addition, to ensure adequately sized strata, we examined the prevalence and incidence of overweight and obesity combined. We tested differences in selected child, maternal, and household characteristics by sex and stunting status at 24 mo with the use of Student's t tests and Pearson's chi-square tests. We compared sex-stratified mean BMI and BMIZ and the prevalence of combined overweight and obesity in each time period by stunting status at 24 mo with the use of Student's t tests and Pearson's chi-square tests, respectively. We compared sex-stratified combined overweight and obesity incidence density rates by stunting status at 24 mo in each time period by calculating incidence density ratios and 95% CIs. Differences in means, frequencies, and rates were considered significant at P < 0.05. We conducted a sensitivity analysis to examine whether cohort participants in the final analytic sample (i.e., those who had length measured at age 2 y and BMI data for all 6 time periods) were significantly different from the individuals who had a length measurement at age 2 y but who were excluded from our analysis because they did not provide BMI data for all 6 time periods. We compared the prevalence of stunting at 24 mo and the prevalence of combined overweight and obesity in each time period by inclusion status with the use of Pearson's chi-square tests. Differences in prevalence were considered significant at P < 0.05. We used SAS version 9.4 (SAS Institute, Inc.) for all analyses. Results The large majority of participants were black (93%). Nearly 26% of males (112 of 438) and 19% of females (88 of 457) in this sample were stunted at 24 mo of age. HAZ and growth patterns over time for this population have been reported elsewhere (29). Birth weight, birth weight z score, HAZ at 24 mo, and maternal height were lower (all P < 0.05) in those who were stunted at 24 mo than in those who were not stunted at 24 mo in both males and females (Table 1). BMI and BMIZ at 24 mo were significantly higher among those who were stunted at 24 mo compared with those who were not stunted at 24 mo in both males and females. The percentage of participants in the lowest-asset quintile did not significantly differ by stunting status at 24 mo in males or females. TABLE 1 Characteristics of South African children by sex and stunting status at 24 mo: Bt20+ cohort1 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 1Values are means ± SDs unless otherwise indicated, n = 895. “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score. View Large TABLE 1 Characteristics of South African children by sex and stunting status at 24 mo: Bt20+ cohort1 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 Males Females Stunted at 24 mo (n = 112) Not stunted at 24 mo (n = 326) P Stunted at 24 mo (n = 88) Not stunted at 24 mo (n = 369) P Child characteristics at birth  Birth weight, g 2930.5 ± 572.5 3194.3 ± 453.8 <0.001 2747.0 ± 506.1 3070.7 ± 444.5 <0.001  Birth weight z score −1.0 ± 1.3 −0.4 ± 1.0 <0.001 −1.2 ± 1.3 −0.4 ± 1.0 <0.001  Gestational age, wk 37.8 ± 1.8 38.1 ± 1.6 0.05 37.5 ± 2.3 38.0 ± 1.8 0.05 Child characteristics at 24 mo  HAZ −2.8 ± 0.7 −0.9 ± 0.7 <0.001 −2.6 ± 0.6 −0.8 ± 0.8 <0.001  BMI, kg/m2 17.0 ± 2.1 16.3 ± 1.8 0.002 16.9 ± 2.2 16.3 ± 1.7 0.007  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.8 ± 1.5 0.4 ± 1.2 0.008 Maternal characteristics  Height, cm 156.8 ± 6.2 159.3 ± 5.8 0.005 156.8 ± 5.7 158.7 ± 6.0 0.01  Education, y 9.4 ± 2.6 9.7 ± 2.5 0.24 9.2 ± 3.0 9.8 ± 2.6 0.09 Household characteristics  Asset quintile 1 (poorest), % 17.4 11.9 0.15 15.9 13.5 0.56 1Values are means ± SDs unless otherwise indicated, n = 895. “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score. View Large Among both males and females, the mean BMI in each of the subsequent time periods did not significantly differ by stunting status at 24 mo (Table 2). Mean BMIZs in early adolescence (11–12 y) and in mid-adolescence (13–15 y) were significantly lower in males who were stunted at 24 mo compared with males who were not stunted at 24 mo (both P < 0.05). Mean BMIZ in early adolescence (11–12 y) was significantly lower among females who were stunted at 24 mo compared with females who were not stunted at 24 mo (P = 0.03). The mean BMIZ at each measurement point by sex and stunting status at 24 mo is shown in Supplemental Figure 2, and the mean BMI at each measurement point by sex and stunting status at 24 mo is shown in Supplemental Figure 3. TABLE 2 BMI, BMIZ, and prevalence of overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 1Values are means ± SDs unless otherwise indicated. ``Overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 2-, 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. View Large TABLE 2 BMI, BMIZ, and prevalence of overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 2 y 4–8 y 11–12 y 13–15 y 16–18 y 22–24 y Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Stunted at 24 mo Not stunted at 24 mo P Males (n = 438, of whom 112 were stunted at 24 mo)  BMI, kg/m2 17.0 ± 12.1 16.3 ± 1.2 0.002 15.5 ± 1.3 15.8 ± 1.3 0.08 17.3 ± 3.2 17.7 ± 3.1 0.14 18.7 ± 3.4 19.2 ± 3.3 0.14 20.0 ± 3.8 20.3 ± 2.9 0.46 21.5 ± 4.5 21.5 ± 3.7 0.99  BMIZ 0.7 ± 1.5 0.2 ± 1.4 0.001 0.0 ± 0.9 0.2 ± 0.9 0.08 −0.5 ± 1.2 0.2 ± 1.2 0.04 −0.7 ± 1.2 0.4 ± 1.2 0.02 −0.7 ± −1.2 0.6 ± 1.0 0.12 N/A N/A  Prevalence overweight and obesity, % 17.0 8.3 0.01 17.0 19.9 0.49 9.9 15.3 0.15 7.2 11.4 0.21 5.4 7.4 0.47 17.0 15.1 0.63 Females (n = 457, of whom 88 were stunted at 24 mo)  BMI, kg/m2 16.9 ± 2.2 16.3 ± 1.7 0.007 15.4 ± 1.4 15.7 ± 1.8 0.08 18.2 ± 3.8 19.1 ± 4.2 0.07 20.6 ± 4.0 21.5 ± 4.5 0.07 22.4 ± 4.2 23.1 ± 4.7 0.22 24.8 ± 6.4 25.9 ± 6.1 0.15  BMIZ 0.8 ± 1.5 0.4 ± 1.2 0.08 −0.1 ± 0.9 0.1 ± 1.0 0.11 −0.3 ± 1.3 0.1 ± 1.3 0.03 0.0 ± 1.3 0.3 ± 1.2 0.06 0.2 ± 1.2 0.4 ± 1.2 0.19 N/A N/A  Prevalence overweight and obesity, % 14.8 9.2 0.12 12.5 14.9 0.56 19.3 21.7 0.63 21.6 27.1 0.29 26.1 28.7 0.63 40.9 49.1 0.17 1Values are means ± SDs unless otherwise indicated. ``Overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 2-, 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. View Large In young adulthood (22–24 y), the prevalence of overweight and obesity was 15.5% in males and 47.5% in females. The increased incidence of overweight and obesity in the females of this cohort has been reported previously (28). Although not widely prevalent among males in any time period, combined overweight and obesity increased over time in females, reaching nearly 41% by young adulthood among those who were stunted at 24 mo and 49% among those who were not stunted at 24 mo. The prevalence of combined overweight and obesity in any of the time periods did not significantly differ by stunting status at 24 mo among either males or females (Table 2). In male participants, the period incidence and incidence density rate of combined overweight and obesity were highest between late adolescence (16–18 y) and young adulthood (22–24 y) (Table 3). These rates did not differ significantly in any time period by stunting status at 24 mo. No males who were stunted at 24 mo became overweight or obese between early adolescence (11–12 y) and mid-adolescence (13–15 y). In females, the period incidence of combined overweight and obesity was highest between late adolescence (16–18 y) and young adulthood (22–24 y), but the incidence density rate was highest between early adolescence (11–12 y) and mid-adolescence (13–15 y). These rates did not differ in any time period by stunting status at 24 mo (Table 3). TABLE 3 Incidence of combined overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 1``Combined overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. 2Incident overweight or obesity cases during the period among those who were at risk of overweight or obesity at the beginning of the period. 3Number of incident overweight or obesity cases during the period per 1000 person-years. 4Rate ratio comparing the incidence of combined overweight and obesity in those who were stunted at 24 mo with those who were not stunted at 24 mo. View Large TABLE 3 Incidence of combined overweight and obesity at selected ages by sex and stunting status at 24 mo: Bt20+ cohort1 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 4–8 to 11–12 y 11–12 to 13–15 y 13–15 to 16–18 y 16–18 to 22–24 y Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Stunted at 24 mo Not stunted at 24 mo Males (n = 438, of whom 112 were stunted at 24 mo)  Period incidence,2 % 5.4 6.4 0 1.2 3.6 1.5 13.4 9.2  Incidence density rate3 8.9 10.8 0 4.9 13.5 5.8 25.7 15.8  Incidence density ratio (95% CI)4 0.8 (0.3, 2.0) N/A 2.3 (0.6, 9.2) 1.6 (0.8, 3.0) Females (n = 457, of whom 88 were stunted at 24 mo)  Period incidence,2 % 11.4 12.5 8.0 9.8 9.1 7.0 19.3 21.7  Incidence density rate3 17.9 20.2 33.5 39.6 33.1 26.7 32.1 38.9  Incidence density ratio (95% CI)4 0.9 (0.4, 1.7) 0.8 (0.3, 1.8) 1.2 (0.5, 2.7) 0.8 (0.5, 1.4) 1``Combined overweight and obesity'' indicates BMIZ >1.0 SD from the WHO reference median for 4- to 8-, 11- to 12-, 13- to 15-, and 16- to 18-y time periods and BMI (kg/m2) ≥25 for the 22- to 24-y time period (23–25). “Stunting” indicates HAZ <–2 SDs of the WHO child growth reference median (22). BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; N/A, not applicable. 2Incident overweight or obesity cases during the period among those who were at risk of overweight or obesity at the beginning of the period. 3Number of incident overweight or obesity cases during the period per 1000 person-years. 4Rate ratio comparing the incidence of combined overweight and obesity in those who were stunted at 24 mo with those who were not stunted at 24 mo. View Large Among participants with a length measurement at age 2 y, the prevalence of stunting at 24 mo in those included in the analyses was not significantly different from that of those who were not included, in either males or females. In males with a length measurement at age 2 y, the prevalence of combined overweight and obesity in mid-adolescence (13–15 y) was significantly different between these 2 groups (P = 0.03) but not in any other time period (Supplemental Table 1). In females with a length measurement at age 2 y, there were no significant differences in the prevalence of combined overweight and obesity between those included in the analyses and those not included in any time period. The distribution of asset quintile significantly differed between participants included in the analyses and participants not included in the analyses (P < 0.001 in both males and females). Between the late-adolescent (16–18 y) and young adult (22–24 y) time periods, 277 participants were lost to follow-up. The prevalence of overweight and obesity did not significantly differ in any earlier time period between participants included in the analyses and participants lost to follow-up after late adolescence. Discussion In this urban South African birth cohort of males and females, we examined the prevalence and incidence of combined overweight and obesity during 5 time periods from early childhood to young adulthood and found no significant differences by stunting status at 24 mo. In LMICs, prospective data are scarce. A strength of our study was the use of longitudinal data spanning the period from early childhood to adulthood. Given the increasing rates of overweight and obesity and related morbidities in LMICs, it is important to understand factors contributing to this trend. Our findings suggest that stunting at 24 mo may not be a significant factor contributing to the increasing overweight and obesity rates in urban South Africa, and therefore that there must be other mechanisms that warrant further research. Previous research has suggested that stunting in childhood may increase the risk of obesity later in life. This association has been shown previously in Russia, China, and Brazil, as well as in South Africa, which are all countries in various stages of nutrition transition (11, 13). An underlying mechanism for this relation is thought to be long-term impaired fat oxidation, a known risk factor for excess weight gain (30, 31). Most studies on this relation have been cross-sectional, and therefore causality has not been established. The relation between childhood stunting and overweight and obesity in South Africa has been examined previously, with mixed findings. Our results at 24 mo affirm previous cross-sectional findings from other transitioning countries, that, at 24 mo, children who were stunted had higher BMI, BMIZ, and prevalence of overweight and obesity than children who were not stunted. However, this association may be spurious and due to random errors in height measurement. Steyn et al. (32), with the use of data from the South African National Food Consumption Survey (NFCS) conducted in 1999, reported that children aged 1–9 y who were short for their age had a higher risk of being overweight. Similarly, Symington et al. (33) with the use of data from the second South African NFCS–Fortification Baseline I conducted in 2005, reported a significant inverse association between BMI and HAZ in South African children aged 3–10 y. On the contrary, Jinabhai et al. (34), using primary data from a community survey in rural South Africa and secondary provincial data from the South African National Primary Schools' Anthropometric (SANPS) Survey in 1994–1995, found no relation between stunting and overweight in children aged 8–11 y, and Mukuddem-Petersen and Kruger (35) found no association between stunting and overweight in schoolchildren aged 10–15 y of age of the North West Province in 2005. A growing body of literature suggests that, when examined longitudinally, there is a null or negative relation between childhood stunting and later overweight and obesity. We extended earlier findings in this same cohort by Cameron and Getz (36), who found no significant relation between stunting status at 24 mo and detailed body composition at 9 y. Outcomes for that study were only collected in late childhood, and the authors suggested that the lack of a relation could be due to the cohort not yet having reached puberty. Our study addressed this gap by using BMI data spanning early childhood to young adulthood. Our findings are consistent with those of Cameron et al. (20) in the childhood time periods. Furthermore, we found no evidence that those who were stunted at 24 mo were any more likely to become overweight or obese in the adolescent and young-adult time periods than those who were not stunted at 24 mo. Our study, however, only used BMI, rather than detailed body-composition data. The findings of our study add to the evidence of a null or negative association between childhood stunting and future overweight and obesity (14–20). Of the longitudinal studies, only 1 study has looked at age-specific patterns of overweight and obesity and found stunting at 6–18 mo to be associated with reduced prevalence of overweight later in childhood in a cohort of Peruvian children (19). The inverse association in that study compared with the null findings in ours could be due to the different ages (6–18 mo compared with 24 mo) used to characterize children as stunted. Like the Peruvian study, we looked at outcomes at numerous follow-up ages; however, a strength of our study was the inclusion of a follow-up point in adulthood. To our knowledge, only 1 other study has prospectively looked at the relation between childhood stunting and adult overweight and obesity (16). Schroeder et al. (16) found an association between childhood stunting and lower BMI and percentage body fat in men, but no association in women in Guatemala. However, after controlling for BMI and percentage body fat, Schroeder et al. found a positive association between stunting and increased abdominal fat in both sexes. The use of BMI in our study to classify overweight and obesity has limitations. BMI is a measure of weight in proportion to height and does not distinguish between lean mass and fat mass. In addition, in using only BMI we are unable to comment on the distribution of subcutaneous fat, which is known to be highly related to morbidity (37). There is evidence to suggest that the trajectory of muscle growth is established prenatally, and that rapid weight gain after undernutrition may lead children to have proportionately less lean muscle tissue and more fat tissue than a child with similar age and weight-for-height who was not undernourished (38, 39). Although we found no association between stunting at age 2 y and BMI in any time period in our study, in a separate subsample of Bt20+ participants, Kagura et al. (40) found an association between stunting at age 2 y and lower lean mass at age 10 y. Future research should aim to examine the relation between stunting in early life and detailed body composition and fat distribution in adulthood. In 2 time periods for males and 1 time period for females, BMIZ significantly differed between those who were stunted at 24 mo and those who were not stunted at 24 mo. However, all differences in BMIZ were in the same direction, and at no point was the BMIZ of those who were stunted at 24 mo higher than the BMIZ of those who were not stunted at 24 mo in either males or females. Differences seen in BMIZ but not BMI were likely due to the process of standardization. By standardizing to a z score we reduced the variability in the measure of BMI. Lower variability increases the ability to detect statistical significance and may have accentuated the differences in BMIZ by stunting status. Our analytical sample was restricted to the 895 participants who had a length measurement at 2 y and a BMI value for each of the 6 time periods. Stratifying the analytic sample by both sex and stunting status resulted in small group sizes, particularly in males, of whom few became overweight or obese in childhood and adolescence. Our analyses may have been underpowered to detect small differences in incidence density within specific age and sex strata. We found no meaningful significant differences in the prevalence of stunting at 24 mo or the prevalence of overweight or obesity in any time period when those with complete longitudinal data were compared with those with incomplete data, leading us to believe that selection bias does not explain our findings. However, our analytic sample had a lower percentage of participants in the poorest-asset quintile compared with those not included in the analyses, so caution must be used when interpreting results. For consistency, we applied the WHO cutoffs for children aged 5–19 y for all childhood and adolescent time periods, which would have overestimated the prevalence of overweight and obesity at age 2 y and, to a lesser extent, at age 4–8 y. However, this would not have affected comparisons between stunted and nonstunted children within any period. Our study adds to the limited evidence on whether childhood stunting is contributing to the increasing rates of overweight and obesity in LMICs and addresses a gap in the literature with regard to follow-up in adolescence and adulthood. We found no evidence that the prevalence or incidence of overweight or obesity in any time period from early childhood to young adulthood differed by stunting at 24 mo in this urban South African cohort. This suggests that stunting may not be an important contributor to the increasing childhood or adult obesity rates in urban South Africa. Acknowledgments The authors’ responsibilities were as follows—SKH, EAL, and ADS: designed the study; SKH: conducted the data analysis with contributions from RJM and EAL; SKH: wrote the first draft of manuscript; LMR and SAN: were the principal investigators of the Birth-to-Twenty Study in South Africa and provided access to the data; RJM, SAN, and ADS: contributed to interpreting the results and writing the final manuscript; and all authors: read and approved the final manuscript. Notes Supported by NIH research training grant R25 TW009337, funded by the Fogarty International Center and the National Institute of Mental Health, and the Department of Science and Technology and National Research Foundation Centre of Excellence in Human Development at the University of the Witwatersrand. The Birth-to-Twenty Study received support from the Wellcome Trust (United Kingdom), the South African Medical Research Council, the University of the Witwatersrand, Johannesburg, the Human Sciences Research Council, and the United Kingdom Medical Research Council/ Department for International Development Africa Research Leader Scheme. Author disclosures: SKH, RJM, EAL, LLM, SAN, and ADS, no conflicts of interest. Supplemental Figures 1–3 and Supplemental Table 1 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn/. Abbreviations used: BMIZ, BMI-for-age z score; Bt20+, Birth-to-Twenty Plus Cohort; HAZ, height-for-age z score; LMIC, low- and middle-income country. References 1. Guh DP , Zhang W , Bansback N , Amarsi Z , Birmingham CL , Anis AH . The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis . BMC Public Health 2009 ; 9 : 88 . Google Scholar CrossRef Search ADS PubMed 2. Ng M , Fleming T , Robinson M , Thomson B , Graetz N , Margono C , Mullany EC , Biryukov S , Abbafati C , Abera SF 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 – 81 . Google Scholar CrossRef Search ADS PubMed 3. Dietz WH . Critical periods in childhood for the development of obesity . Am J Clin Nutr 1994 ; 59 : 955 – 9 . Google Scholar CrossRef Search ADS PubMed 4. de Onis M , Blossner M , Borghi E . Global prevalence and trends of overweight and obesity among preschool children . Am J Clin Nutr 2010 ; 92 : 1257 – 64 . Google Scholar CrossRef Search ADS PubMed 5. de Onis M , Blossner M , Borghi E . Prevalence and trends of stunting among pre-school children, 1990–2020 . Public Health Nutr 2012 ; 15 : 142 – 8 . Google Scholar CrossRef Search ADS PubMed 6. Tzioumis E , Adair LS . Childhood dual burden of under- and overnutrition in low- and middle-income countries: a critical review . Food Nutr Bull 2014 ; 35 : 230 – 43 . Google Scholar CrossRef Search ADS PubMed 7. Popkin BM . The nutrition transition and its health implications in lower-income countries . Public Health Nutr 1998 ; 1 : 5 – 21 . Google Scholar CrossRef Search ADS PubMed 8. Said-Mohamed R , Micklesfield LK , Pettifor JM , Norris SA . Has the prevalence of stunting in South African children changed in 40 years? A systematic review . BMC Public Health 2015 ; 15 : 534 . Google Scholar CrossRef Search ADS PubMed 9. Pettifor JM . Combined stunting and overweight in young children—a paradox? South Afr J Clin Nutr 2006 ; 19 : 98 – 100 . Google Scholar CrossRef Search ADS PubMed 10. Bourne LT , Lambert EV , Steyn K . Where does the black population of South Africa stand on the nutrition transition? Public Health Nutr 2002 ; 5 : 157 – 62 . Google Scholar CrossRef Search ADS PubMed 11. Popkin BM , Richards MK , Montiero CA . Stunting is associated with overweight in children of four nations that are undergoing the nutrition transition . J Nutr 1996 ; 126 : 3009 – 16 . Google Scholar CrossRef Search ADS PubMed 12. Barker DJ . The fetal and infant origins of adult disease . BMJ 1990 ; 301 : 1111 . Google Scholar CrossRef Search ADS PubMed 13. Sawaya AL , Dallal G , Solymos G , de Sousa MH , Ventura ML , Roberts SB , Sigulem DM . Obesity and malnutrition in a Shantytown population in the city of Sao Paulo, Brazil . Obes Res 1995 ; 3 ( Suppl 2 ): 107s – 15s . Google Scholar CrossRef Search ADS PubMed 14. Walker SP , Gaskin PS , Powell CA , Bennett FI . The effects of birth weight and postnatal linear growth retardation on body mass index, fatness and fat distribution in mid and late childhood . Public Health Nutr 2002 ; 5 : 391 – 6 . Google Scholar CrossRef Search ADS PubMed 15. Walker SP , Chang SM , Powell CA . The association between early childhood stunting and weight status in late adolescence . Int J Obes (Lond) 2007 ; 31 : 347 – 52 . Google Scholar CrossRef Search ADS PubMed 16. Schroeder DG , Martorell R , Flores R . Infant and child growth and fatness and fat distribution in Guatemalan adults . Am J Epidemiol 1999 ; 149 : 177 – 85 . Google Scholar CrossRef Search ADS PubMed 17. Benefice E , Garnier D , Simondon KB , Malina RM . Relationship between stunting in infancy and growth and fat distribution during adolescence in Senegalese girls . Eur J Clin Nutr 2001 ; 55 : 50 – 58 . Google Scholar CrossRef Search ADS PubMed 18. Tanner S , Leonard WR , Reyes-Garcia V . The consequences of linear growth stunting: influence on body composition among youth in the Bolivian Amazon . Am J Phys Anthropol 2014 ; 153 : 92 – 102 . Google Scholar CrossRef Search ADS PubMed 19. Andersen CT , Stein AD , Reynolds SA , Behrman JR , Crookston BT , Dearden KA , Penny ME , Schott W , Fernald LCH . Stunting in infancy is associated with decreased risk of high body mass index for age at 8 and 12 years of age . J Nutr 2016 ; 146 : 2296 – 303 . Google Scholar CrossRef Search ADS PubMed 20. Cameron N , Wright MM , Griffiths PL , Norris SA , Pettifor JM . Stunting at 2 years in relation to body composition at 9 years in African urban children . Obes Res 2005 ; 13 : 131 – 6 . Google Scholar CrossRef Search ADS PubMed 21. Richter L , Norris S , Pettifor J , Yach D , Cameron N . Cohort profile: Mandela's children: the 1990 Birth to Twenty Study in South Africa . Int J Epidemiol 2007 ; 36 : 504 – 11 . Google Scholar CrossRef Search ADS PubMed 22. WHO Multicentre Growth Reference Study Group . WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development . Geneva (Switzerland) : WHO ; 2006 . 23. de Onis M , Onyango AW , Borghi E , Siyam A , Nishida C , Siekmann J . Development of a WHO growth reference for school-aged children and adolescents . Bull World Health Organ 2007 ; 85 : 660 – 7 . Google Scholar CrossRef Search ADS PubMed 24. Cole TJ , Bellizzi MC , Flegal KM , Dietz WH . Establishing a standard definition for child overweight and obesity worldwide: international survey . BMJ 2000 ; 320 : 1240 – 3 . Google Scholar CrossRef Search ADS PubMed 25. WHO . Obesity: preventing and managing the global epidemic . Report of a WHO consultation . World Health Organ Tech Rep Ser 2000 ; 894 : i – xii , 1–253 . PubMed 26. Van den Broeck J , Willie D , Younger N . The World Health Organization child growth standards: expected implications for clinical and epidemiological research . Eur J Pediatr 2009 ; 168 : 247 – 51 . Google Scholar CrossRef Search ADS PubMed 27. de Onis M , Habicht JP . Anthropometric reference data for international use: recommendations from a World Health Organization Expert Committee . Am J Clin Nutr 1996 ; 64 : 650 – 8 . Google Scholar CrossRef Search ADS PubMed 28. Lundeen EA , Norris SA , Adair LS , Richter LM , Stein AD . Sex differences in obesity incidence: 20-year prospective cohort in South Africa . Pediatr Obes 2016 ; 11 ( 1 ): 75 – 80 . Google Scholar CrossRef Search ADS PubMed 29. Stein AD , Wang M , Martorell R , Norris SA , Adair LS , Bas I , Sachdev HS , Bhargava SK , Fall CH , Gigante DP et al. Growth patterns in early childhood and final attained stature: data from five birth cohorts from low- and middle-income countries . Am J Hum Biol 2010 ; 22 : 353 – 9 . Google Scholar CrossRef Search ADS PubMed 30. Hoffman DJ , Sawaya AL , Verreschi I , Tucker KL , Roberts SB . Why are nutritionally stunted children at increased risk of obesity? Studies of metabolic rate and fat oxidation in shantytown children from Sao Paulo, Brazil . Am J Clin Nutr 2000 ; 72 : 702 – 7 . Google Scholar CrossRef Search ADS PubMed 31. Sawaya AL , Grillo LP , Verreschi I , da Silva AC , Roberts SB . Mild stunting is associated with higher susceptibility to the effects of high fat diets: studies in a shantytown population in Sao Paulo, Brazil . J Nutr 1998 ; 128 ( Suppl ): 415s – 20s . Google Scholar CrossRef Search ADS PubMed 32. Steyn NP , Labadarios D , Maunder E , Nel J , Lombard C . Secondary anthropometric data analysis of the National Food Consumption Survey in South Africa: the double burden . Nutrition 2005 ; 21 : 4 – 13 . Google Scholar CrossRef Search ADS PubMed 33. Symington EA , Gericke GJ , Nel JH , Labadarios D . The relationship between stunting and overweight among children from South Africa: secondary analysis of the National Food Consumption Survey–Fortification Baseline I . S Afr Med J 2015 ; 106 : 65 – 69 . Google Scholar CrossRef Search ADS PubMed 34. Jinabhai CC , Taylor M , Sullivan KR . Implications of the prevalence of stunting, overweight and obesity amongst South African primary school children: a possible nutritional transition? Eur J Clin Nutr 2003 ; 57 : 358 – 65 . Google Scholar CrossRef Search ADS PubMed 35. Mukuddem-Petersen J , Kruger HS . Association between stunting and overweight among 10-15-y-old children in the North West Province of South Africa: the THUSA BANA study . Int J Obes Relat Metab Disord 2004 ; 28 : 842 – 51 . Google Scholar CrossRef Search ADS PubMed 36. Cameron N , Getz B . Sex differences in the prevalence of obesity in rural African adolescents . Int J Obes Relat Metab Disord 1997 ; 21 : 775 – 82 . Google Scholar CrossRef Search ADS PubMed 37. Botton J , Heude B , Kettaneh A , Borys JM , Lommez A , Bresson JL , Ducimetiere P , Charles MA . Cardiovascular risk factor levels and their relationships with overweight and fat distribution in children: the Fleurbaix Laventie Ville Sante II Study . Metabolism 2007 ; 56 : 614 – 22 . Google Scholar CrossRef Search ADS PubMed 38. Singhal A , Wells J , Cole TJ , Fewtrell M , Lucas A . Programming of lean body mass: a link between birth weight, obesity, and cardiovascular disease? Am J Clin Nutr 2003 ; 77 : 726 – 30 . Google Scholar CrossRef Search ADS PubMed 39. Sawaya AL , Roberts S . Stunting and future risk of obesity: principal physiological mechanisms . Cad Saude Publica 2003 ; 19 ( Suppl 1 ): S21 – 8 . Google Scholar CrossRef Search ADS PubMed 40. Kagura J , Feeley AB , Micklesfield LK , Pettifor JM , Norris SA . Association between infant nutrition and anthropometry, and pre-pubertal body composition in urban South African children . J Dev Orig Health Dis 2012 ; 3 : 415 – 23 . Google Scholar CrossRef Search ADS PubMed © 2018 American Society for Nutrition. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of NutritionOxford University Press

Published: May 15, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off