Anthropometric Characterization of Impaired Fetal Growth: Risk Factors for and Prognosis of Newborns With Stunting or Wasting

Importance  Stunting (short length for age) and wasting (low body mass index [BMI] for age) are widely used to assess child nutrition. In contrast, newborns tend to be assessed solely based on their weight.

Objective  To use recent international standards for newborn size by gestational age to assess how stunted and wasted newborns differ in terms of risk factors and prognoses.

Design, Setting, and Participants  A cross-sectional study with follow-up until hospital discharge was conducted at urban sites in Brazil, China, India, Italy, Kenya, Oman, England, and the United States that are participating in the INTERGROWTH-21^st Project. The study was conducted from April 27, 2009, to March 2, 2014, and the final dataset for analyses was locked on March 19, 2014.

Exposures  Sociodemographic and behavioral maternal risk factors, previous pregnancy history, and maternal and fetal conditions during pregnancy were investigated as risk factors for stunting and wasting. Anthropometry at birth was used to predict for neonatal prognosis.

Main Outcomes and Measures  Newborn stunting and wasting were defined as birth length and BMI for gestational age below the third centiles of the INTERGROWTH-21^st standards. Prognosis was assessed through mortality before hospital discharge, admission to neonatal intensive care units, and newborn complications.

Results  From the 60 206 singleton live births during the study period, we selected all newborns between 33 weeks’ and 42 weeks 6 days’ gestation at birth (51 200 [85%]) with reliable ultrasound dating. Stunting affected 3.8% and wasting 3.4% of all newborns; both conditions were present in 0.7% of the sample. Of the 26 conditions studied, five were more strongly associated with stunting than with wasting (reported as odds ratios [OR]; 95% CI): short maternal height (6.7; 5.1-9.0), younger maternal age (0.7; 0.5-0.9), smoking (2.8; 2.3-3.3), illicit drug use (2.3; 1.5-3.6), and clinically suspected intrauterine growth restriction (5.2; 4.5-6.0). Wasting was more strongly related than stunting with 4 newborn outcomes (neonatal intensive care stay, 6.7 [5.5-8.1]; respiratory distress syndrome, 4.0 [3.3-4.9]; transient tachypnea, 2.1 [1.5-2.9]; and no oral feeding for >24 hours, 5.0 [3.9-6.5]). Maternal gestational diabetes mellitus was protective against wasting (0.6; 0.5-0.8) but not against stunting (0.9; 0.7-1.1).

Conclusions and Relevance  Although newborn stunting and wasting share some common determinants, they are distinct phenotypes with their own risk factors and neonatal prognoses. To be consistent with the literature on infant and child nutrition, newborns should be classified using the 2 phenotypes of stunting and wasting. The distinction will help to prioritize preventive interventions and focus the management of fetal undernutrition.

Neonatologists have consistently differentiated proportionately from disproportionately developed newborns using a combination of anthropometric measures, such as Rohrer’s Ponderal Index¹ (weight/length²). Curves relating the Ponderal Index to gestational age at birth (based on babies born between 1948 and 1961 in a US hospital³) are widely used for clinical and research purposes, but, in many countries, birth weight alone is used.

This lack of a consensus conflicts with the literature² on infant and child nutrition that recognizes that low weight at a given age may result from stunting (short length for age, reflecting linear growth restriction), wasting (low weight for length, or low body mass index [BMI] for age, often reflecting recent weight loss), or both these phenotypes. This body of literature acknowledges that although infant and child stunting and wasting share some common determinants, they are fundamentally 2 distinct phenotypes with different timing and duration of the causal insults, specific risk factors, varied distributions across populations, and different prognoses. The reasons for the disparities in evaluating and classifying the nutritional status of newborns and infants across such a short period of life remain unclear.

Classifying newborns as proportionately or disproportionately developed⁴ and infants as stunted or wasted is conceptually similar, but different criteria are used to define these phenotypes. Therefore, we used the recently published International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21^st) Project prescriptive standards for newborn length and weight⁵ to produce corresponding BMI values (calculated as weight in kilograms divided by length in meters squared) for gestational age standards, enabling newborns to be characterized using the same indicators already applied in infancy and childhood.

In the present study, we used the INTERGROWTH-21^st Project length and BMI for gestational age standards to classify more than 50 000 newborns. We tested the hypothesis that stunting and wasting evaluated at birth are separate phenotypes that, despite sharing some common risk factors, have different determinants and prognostic implications.

Box Section Ref ID

Newborn size was measured in the Newborn Cross Sectional Study, a component of the INTERGROWTH-21^st Project. The birth weight and birth length of newborn babies from 8 geographically defined urban populations were measured using the same methods and identical equipment. A detailed description of the study design, methods, and strategy for selecting the study populations is available elsewhere.⁶

The INTERGROWTH-21^st Project was a multicenter, multiethnic, population-based study conducted between April 27, 2009, and March 2, 2014, in Pelotas, Brazil; Turin, Italy; Muscat, Oman; Oxford, England; Seattle, Washington; Shunyi County, a suburban district of Beijing, China; Central Nagpur, India; and Parklands suburb, Nairobi, Kenya. Data analysis was conducted on the final data set that was locked on March 19, 2014. The INTERGROWTH-21^st Project’s protocol was approved by the Oxfordshire Research Ethics Committee C, as well as the ethics committee of each institution and health authority where the project was implemented. In the Newborn Cross Sectional Study, we obtained institutional consent to use routinely collected data and the women gave verbal consent. The participants received no financial compensation.

Participating hospitals covered more than 80% of all deliveries in their corresponding geographically demarcated areas. Data collection continued for 12 consecutive months at each site or until the target of more than 7000 deliveries per site was attained. The strategy of studying the entire unselected newborn population allowed a large number of small-for-gestational-age and preterm babies to be identified.

All singleton newborns with a reliable gestational age validated by ultrasonography between 33 and 42 weeks’ gestation, as defined below, were included in the present analysis. This population constituted the Newborn Cross-Sectional Study, 1 of the 3 main components of the INTERGROWTH-21^st Project.

During the preparatory phase of the INTERGROWTH-21^st Project, all participating hospitals adopted a policy of estimating gestational age at the first antenatal visit by performing an ultrasonographic examination to measure either fetal crown to rump length (if <14 weeks’ gestation) or head circumference (if ≤24 weeks’ gestation). If the pregnancy was more than 24 weeks’ gestation, the estimate was considered reliable only if it was within 1 week of the gestational age estimation based on the date of the woman’s last menstrual period.⁷

Information on maternal social, demographic, environmental, and clinical characteristics, as well as pregnancy and delivery outcomes, was taken from the medical records and complemented by information from health care professionals (if records were incomplete) and interviewing mothers using a structured questionnaire.

Several risk factors that could affect newborn body size were studied. Social and biological maternal variables comprised maternal educational level in completed years, age, parity, height, weight, and BMI. Maternal behavior during pregnancy included reported smoking and use of alcohol or illicit drugs. For parous women, reproductive history was collected on previous low-birth-weight and preterm newborns, stillbirths, and newborn deaths. Maternal conditions during the index pregnancy comprised epilepsy, gestational diabetes mellitus, pregnancy-induced hypertension, preeclampsia, severe preeclampsia or eclampsia, and suspected intrauterine growth restriction defined by ultrasonographic evidence of growth restriction or a clinical indication reported in the medical records.

Newborn anthropometric measures were obtained within 12 hours of birth, using identical equipment at all sites: electronic scale for birth weight (Seca), recumbent length using a specially designed Harpenden infantometer (Chasmors Ltd), and head circumference using a metallic nonextendable tape measure (Chasmors Ltd).⁸ Measurement procedures and protocols were standardized based on World Health Organization recommendations to ensure maximum validity.⁹

The body measurements were collected independently in duplicate by 2 trained anthropometrists for all births, including stillbirths. If differences between measurements exceeded the set maximum allowable values (birth weight, 50 g; birth length, 7 mm; and head circumference, 5 mm), both observers independently obtained a second measurement. The intraobserver and interobserver error of measurement values, obtained during the standardization and retraining sessions of anthropometry staff, were 0.3 to 0.5 cm for recumbent length.¹⁰

During the study period, all newborns, including those admitted to the neonatal intensive care unit (NICU) at the special care level or another referral care level, were assessed daily until hospital discharge. We captured 3 newborn outcomes: (1) death before hospital discharge, (2) NICU admission for 7 or more days (as a proxy for severe neonatal morbidity), and (3) newborn morbidities affecting more than 1% of the sample (ie, respiratory distress syndrome, transitory tachypnea, neonatal sepsis, hyperbilirubinemia, hypoglycemia, no oral feeding for >24 hours⁷). Neonatal outcomes were assessed until hospital discharge.

Methods for training, standardization, and quality control were uniformly used across all sites.¹⁰ Neonatal clinical practices, including NICU care and feeding, were also standardized based on a package of minimum evidence-based practices following a protocol adopted by the INTERGROWTH-21^st Project Neonatal Study Group.^10,11

All supporting documentation and data collection forms were prepared by the Project Coordinating Unit, translated into the main local language, tested locally, and introduced into our electronic data management system (http://www.intergrowth21.org.uk).¹² All forms were linked to reduce duplication and facilitate quality control measures.

Newborns were divided into 2 distinct impaired fetal growth phenotypes (stunted and wasted) based on birth length and BMI measurements at birth that were less than the respective third centiles of the INTERGROWTH-21^st Project Newborn Size Standards.^13,14 A newborn could also be classified as both stunted and wasted.

In separate analyses, we investigated risk factors associated with newborn stunting and wasting and then the neonatal prognoses for these phenotypes. Published risk factors on low birth weight were explored using multivariable logistic regression models. Six maternal variables (educational level, height, weight, BMI, age, and parity) were treated as categorical variables (as presented in the Results section); the remaining risk factors and outcomes were dichotomous.

A conceptual model was defined a priori to guide the analyses.¹⁵ Distal maternal determinants, including the mother’s social and biological characteristics, and behaviors (smoking, as well as alcohol or illicit drug use) were adjusted for one another. Proximate determinants (ie, maternal and fetal conditions diagnosed during the index pregnancy) were adjusted for one another and for the distal maternal determinants that represent potential confounders.¹⁵

Previous pregnancy outcomes (low birth weight, preterm birth, fetal or neonatal deaths) were deemed to be risk markers rather than determinants and were not included as confounders in these models. All regression analyses included the study site as a categorical covariate. In the first set of logistic models, fetal stunting and wasting were used as separate explanatory variables for newborn mortality and morbidity, as well as for NICU stay. The second set of analyses assessed the roles of the above distal and proximate determinants on newborn stunting and wasting.

The results were adjusted for study site and possible confounders (as described above). Unless otherwise stated all associations reported were significant at P < .05. Bootstrapping procedures with 1000 independent samples were used for each risk factor and prognostic outcome to compare the log odds ratios (ORs) associated with stunting and wasting. Statistical analyses were carried out with IBM SPSS Statistics, version 22 (IBM Corp) and Stata, version 11 (StataCorp LP).

The proportional contributions of newborns to the total population by site were Brazil (9.8%), China (12.4%), India (13.2%), Italy (12.8%), Kenya (12.8%), Oman (13.8%), England (14.5%), and the United States (10.6%). The total sample included 24 817 girls (48.5%) and 26 378 boys (51.5%); information was missing on sex for 5 newborns. Information on birth weight was missing for 114 newborns (0.2%) and on birth length for 894 (1.7%) of all newborns.

From the 60 206 singleton live births during the study period, which ended on March 2, 2014, we selected all newborns between 33 weeks’ and 42 weeks 6 days’ gestation at birth (51 200 [85%]) with reliable ultrasonographic dating.

Stunting and wasting affected 3.8% (1944 of 51 086) and 3.4% (1729 of 50 306) of all newborns, respectively; 0.7% of the newborns (n = 344) were stunted and wasted. The mean gestational ages of the 4 newborn subgroups (eTable in the Supplement) were similar. Those who were stunted and wasted were lighter than the rest of the sample, with smaller head circumferences. Stunted-only newborns were 120 g heavier than wasted-only newborns but otherwise similar. We then compared all stunted with nonstunted (including wasted) newborns, and all wasted with nonwasted newborns.

Tables 1, 2, 3, and 4 report the associations between newborn stunting and wasting with risk and prognostic factors. The Box presents the results of the analyses testing whether, for a given risk factor or prognosis, there was a significant difference between the ORs for stunting and wasting.

Box Section Ref ID

A total of 211 newborns (mortality rate of 4 per 1000) died before hospital discharge, and 1556 newborns (3.0%) were admitted to the NICU for 7 or more days. After adjusting for study site and maternal social and biological confounders, the ORs (95% CIs) for neonatal mortality associated with stunting and wasting were 7.4 (3.9-14.4) and 3.7 (1.5-9.6), respectively. In the last column in Table 1, we adjusted newborn BMI for length (both were adjusted for gestational age) and vice versa, including them as continuous variables in the models. In these adjusted analyses, stunting remained significantly associated with neonatal mortality; wasting did not.

Newborn wasting was more strongly associated than stunting with poor neonatal outcomes, including NICU stay of 7 days or more, respiratory distress syndrome, transient tachypnea, and no oral feeding for more than 24 hours (Table 1 and Box). Results for hyperbilirubinemia and hypoglycemia in the crude analyses, as observed in the column showing the prevalence of conditions, were substantially different from the adjusted results because of the heterogeneity in the reported prevalence of these conditions across study sites.

After demonstrating that the 2 phenotypes have distinct neonatal outcomes, including specific morbidity profiles, we conducted multivariable analyses to identify maternal and pregnancy characteristics differentially associated with stunting and wasting. Table 2 reports the association between the 2 phenotypes and maternal social and biological risk factors. Low educational level (<8 years’ schooling) was associated with a 20% to 40% greater risk of stunting and wasting compared with higher educational levels (≥12 years). The newborns of short mothers (<150 cm) were 6.7 times more likely to be stunted but 1.9 times more likely to be wasted compared with those who were taller (≥170 cm). Low maternal BMI (<18.5) was associated with an approximately 2-fold increase in risk of both stunting and wasting compared with a BMI of 25 or more. Associations with maternal age were not as clear cut, and 95% CIs often included unity.

In the unadjusted analyses, the lowest risk was for newborns of women aged 30 to 35 years; however, after adjustment for confounding, young mothers were at the lowest risk. Parity was inversely related to stunting, with a 2-fold increase for primiparae compared with mothers with parity of 3 or more; the association with wasting was less marked. Maternal smoking was associated with ORs of 2.8 for stunting and 1.5 for wasting; for illicit drug use, the corresponding ORs were 2.3 and 0.6. Alcohol use was not associated with either outcome in the confounder-adjusted analyses. Among all the risk factors listed in Table 2, maternal height, age, smoking, and illicit drug use were more strongly associated with stunting than with wasting (Box).

The analyses in Table 3 are limited to 24 192 women with a previous delivery. For those with a previous low-birth-weight or preterm newborn, the adjusted ORs were 2.0 and 1.8 for stunting and 2.5 and 1.6 for wasting, respectively. A previous stillbirth or neonatal death was not associated with stunting but showed a marginal association with wasting. None of these factors was more strongly associated with stunting than wasting, or vice versa (Box). Because previous obstetric history is a risk marker rather than a potential determinant of fetal growth, these variables were not included as confounders in the next sets of analyses.

Table 4 describes the results for selected maternal clinical conditions during pregnancy. Epilepsy was associated with a nonsignificant doubling of the frequency of both outcomes, but only 87 mothers had this diagnosis. Gestational diabetes was associated with a lower OR (95% CI) for wasting (0.6; 0.5-0.8), but not for stunting (0.9; 0.7-1.1).

Pregnancy-related hypertension, preeclampsia, and severe preeclampsia or eclampsia all had slightly stronger associations with wasting than with stunting. All associations were statistically significant, but ORs were largest for severe preeclampsia or eclampsia (2.5 vs 2.2) compared with preeclampsia (1.9 vs 1.4) and smallest for pregnancy-related hypertension (1.6 vs 1.3), respectively (Table 4).

Suspected intrauterine growth restriction, defined based on ultrasonographic evidence or if specifically mentioned in the medical records, was more strongly associated with stunting (OR, 5.2) than with wasting (OR, 3.2). TORCH (toxoplasmosis, other [syphilis, varicella zoster, parvovirus B19], rubella, cytomegalovirus, and herpes infections) was associated with both outcomes, with ORs of 3.7 for stunting and 2.7 for wasting, and severe congenital malformations were associated with a doubling in the frequencies of both stunting (OR, 2.3) and wasting (OR, 2.2).

Of the 26 conditions studied (Box), short maternal height, younger maternal age, smoking, gestational diabetes, drug use, and suspected intrauterine growth restriction were more strongly associated with stunting than with wasting. However, wasting was more strongly related than stunting with 4 newborn outcomes (NICU stay, respiratory distress syndrome, transient tachypnea, and no oral feeding for >24 hours). Gestational diabetes was protective against wasting but not stunting. For the remaining 16 conditions, there was no statistically significant evidence of a differential association with either stunting or wasting.

We have provided considerable clinical and epidemiologic evidence from the Newborn Cross-Sectional Study component of the population-based INTERGROWTH-21^st Project to support the concept that stunting and wasting are separate anthropometric phenotypes with intrauterine origins. Although not unexpected, since growth and development from conception to childhood is a biological continuum, the findings improve our understanding of fetal growth alterations beyond the simple evaluation of birth weight.

Our analyses’ strengths include the large multicountry sample size and the strict pregnancy dating and neonatal standardization procedures across the study sites. In addition, use of the recently published International Newborn Size for Gestational Age standards⁵ allowed more precise classification of anthropometric deficits than previously possible. Our analyses focused first on demonstrating that stunted and wasted newborns differ in terms of short-term prognosis. Having shown this, we assessed how the phenotypes differed in terms of risk factors. As expected, some conditions were associated with stunting and wasting with similar strength, mostly those recognized as universal risk factors (eg, educational level, maternal undernutrition, obstetric history). Other factors, in particular the conditions from mild preeclampsia to eclampsia, have such a wide range of severities, presentations, and timing during pregnancy that they are not phenotype specific.

A limitation of our analyses is that the Newborn Cross-Sectional Study follow-up ended at hospital discharge, precluding the analyses of associations between newborn stunting and wasting with longer-term outcomes. However, we are assessing growth and development in a long-term follow-up study of the total population enrolled in the Fetal Growth Longitudinal Study of the INTERGROWTH-21^st Project that may provide an even better understanding of the phenotypes.

The observed prevalences of newborn stunting and wasting were relatively low probably because the Newborn Cross-Sectional Study was performed in generally low-risk populations that were identified as appropriate for selecting mothers to include in the Fetal Growth Standards. Our analyses used Newborn Size Standards derived from a subsample of approximately 20 000 women from the more than 50 000 women included in the present analyses.⁵ The subsample was restricted to women who had early ultrasonographic measurement and who met strict individual eligibility criteria for those at low risk of fetal growth impairment. In addition, stunting at birth seems to have a relatively low prevalence even in low-income settings, with a prevalence increasing sharply with age.¹⁶ In future studies, populations with a higher prevalence of the risk factors analyzed or with different risk factors (eg, malaria, smoking) may strengthen the associations with the 2 phenotypes and almost certainly reinforce the differences between stunting and wasting.

We opted not to correct statistical significance levels for multiple testing. The associations studied are based on a priori hypotheses, because the risk factors and prognostic outcomes are knowingly associated with low birth weight. Our goal was to discover whether these associations are related to the wasting or stunting components of low weight. In addition, we reported exact significance levels so that readers may verify the probability level of each association.

In terms of the association with neonatal outcomes (Table 1), wasted newborns were more likely to develop neonatal complications than were stunted newborns, perhaps because those with wasting have less fat deposits. However, in terms of mortality prediction, the ORs for stunting were seemingly higher than were those for wasting, but both ORs had wide 95% CIs, and there was no statistically significant difference (Box). In the pediatric literature,¹⁷ severe wasting is associated with a higher risk for mortality than is severe stunting, but the risks are similar for both phenotypes when moderate cases were considered.

Neonatal stunting could be related to organic conditions (including undiagnosed malformations) that are less treatable by NICU interventions. For every morbidity, but not for overall mortality, the prognosis for wasted newborns tended to be worse than for stunted newborns. Findings from the analyses of social and biological risk factors confirmed several well-described associations in the birth weight literature; examples of these include maternal educational level, weight, height, smoking, and parity.¹⁸ Our results show a specific differential association between maternal height and stunting but a weaker association with wasting.

Our findings on the association of stunting and wasting with maternal age show the importance of adjustment for social and biological confounders; an apparent increase in risk for children born to young mothers was inverted after such adjustment.¹⁹ Maternal smoking and reported use of illicit drugs were more strongly associated with stunting than with wasting. Because these were prepregnancy habits, this effect may be related to exposure from oocyte maturation to birth.

Gestational diabetes was associated with a reduced risk of wasting (a well-described effect that is due to increased birth weight and fat deposition), but it had no association with stunting. There was also an expected gradient in both stunting and wasting for the blood pressure–related conditions: the more severe the condition, the stronger the association with both wasting and stunting. Conversely, the association with suspected intrauterine growth restriction was stronger for stunting than for wasting.

Fetal stunting is widely regarded as a cumulative, “long-term” process analogous to chronic undernutrition in children² that requires exposure to 1 or more risk factors for several months or throughout pregnancy; neonatal wasting is likely to reflect acute exposures in the weeks before delivery when fat deposition occurs more rapidly.⁴ Other investigators,²⁰ however, have postulated that differences in severity, rather than the timing and duration of the insults, result in the distinct phenotypes of impaired fetal growth, with wasting representing the more severe cases.

Impaired fetal growth (for which impaired newborn size is a proxy) is a complex syndrome. Its components are interlinked with multiple etiologic factors that have specific short- and long-term effects. The proportional distribution of impaired fetal growth across populations is dependent on the prevalence of multiple, underlying causal factors. Further characterization of phenotypes and validation in different populations is needed, as we have attempted for the preterm delivery syndrome.²¹

We envisage that, in time, the phenotypic characterization of growth disturbances described in this study could, after testing across populations, be refined by research findings into fetal organ growth and placental blood flow patterns as well as biological markers related to the etiology or pathophysiology of the causal conditions. The markers could include maternal nutrient or growth factor levels, immunologic or inflammatory markers, epigenetic profiles, and genetic variants. The phenotypic characterization of growth disturbances should extend to the other end of the spectrum to include overweight disorders, which represent a major public health problem across the world.

Diminished size at birth and suboptimal fetal growth patterns, particularly when used in conjunction with gestational age, are powerful predictors of short-term morbidity and mortality, developmental outcomes, long-term health, and the risk of chronic disease in adulthood.^22-24 We have shown that classification of newborns in terms of both stunting and wasting improves the understanding of risk factors and prognosis. In pediatric practice, sole reliance on weight for age is being replaced by assessments of linear growth (length) and adiposity (BMI). We propose that it is time to incorporate these concepts into the anthropometric assessment of newborns.

Accepted for Publication: May 7, 2015.

Corresponding Author: José Villar, MD, Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Women’s Centre, Level 3, John Radcliffe Hospital, Headington, Oxford OX3 9DU, England (jose.villar@obs-gyn.ox.ac.uk).

Published Online: July 6, 2015. doi:10.1001/jamapediatrics.2015.1431.

Author Contributions: Professors Villar, Bhutta, and Kennedy contributed equally to the study. Professors Victora and Villar had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Victora, Villar, Barros, Ismail, Chumlea, Bertino, Jaffer, Noble, Pang, Bhutta, Kennedy.

Acquisition, analysis, or interpretation of data: Victora, Villar, Barros, Ismail, Chumlea, Papageorghiou, Ohuma, Lambert, Carvalho, Altman, Noble, Gravett, Purwar, Frederick.

Drafting of the manuscript: Victora, Villar, Barros, Purwar, Kennedy.

Critical revision of the manuscript for important intellectual content: Villar, Barros, Ismail, Chumlea, Papageorghiou, Bertino, Ohuma, Lambert, Carvalho, Jaffer, Altman, Noble, Noble, Gravett, Frederick, Pang, Bhutta, Kennedy.

Statistical analysis: Victora, Villar, Barros, Ohuma, Altman.

Obtained funding: Villar, Kennedy.

Administrative, technical, or material support: Barros, Ismail, Papageorghiou, Lambert, Jaffer, Noble, Gravett, Purwar, Pang, Bhutta, Kennedy.

Study supervision: Barros, Ismail, Chumlea, Papageorghiou, Bertino, Carvalho, Jaffer, Purwar.

Conflict of Interest Disclosures: None reported.

Funding/Support: This project was supported by a generous grant (49038) from the Bill & Melinda Gates Foundation to the University of Oxford, for which we are very grateful. Philips Medical Systems provided the ultrasonography equipment and technical assistance throughout the project. MedSciNet UK Ltd set up the INTERGROWTH-21^st Project website and developed, maintained, and supported the online data management system.

Role of the Funder/Sponsor: The funding sources played no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Group Information: Members of the International Fetal and Newborn Growth Consortium for the 21st Century (INTERGROWTH-21^st) and its committees are listed in the eAppendix in the Supplement.

Additional Contributions: We thank the health authorities in Pelotas, Brazil; Beijing, China; Nagpur, India; Turin, Italy; Nairobi, Kenya; Muscat, Oman; Oxford, England; and Seattle, Washington, who facilitated the project by allowing participation of these study sites as collaborating centers. We thank the parents and infants who participated in the studies and the more than 200 members of the research teams who made the implementation of this project possible. The participating sites included: Pelotas, Brazil (Hospital Miguel Piltcher, Hospital São Francisco de Paula, Santa Casa de Misericórdia de Pelotas, and Hospital Escola da Universidade Federal de Pelotas); Beijing, China (Beijing Obstetrics and Gynecology Hospital, Shunyi Maternal and Child Health Centre, and Shunyi General Hospital); Nagpur, India (Ketkar Hospital, Avanti Institute of Cardiology Private Limited, Avantika Hospital, Gurukrupa Maternity Hospital, Mulik Hospital and Research Centre, Nandlok Hospital, Om Women’s Hospital, Renuka Hospital and Maternity Home, Saboo Hospital, Brajmonhan Taori Memorial Hospital, and Somani Nursing Home); Nairobi, Kenya (Aga Khan University Hospital, MP Shah Hospital, and Avenue Hospital); Turin, Italy (Ospedale Infantile Regina Margherita Sant’ Anna and Azienda Ospedaliera Ordine Mauriziano); Muscat, Oman (Khoula Hospital, Royal Hospital, Wattayah Obstetrics and Gynaecology Poly Clinic, Wattayah Health Centre, Ruwi Health Centre, Al-Ghoubra Health Centre, and Al-Khuwair Health Centre); Oxford, England (John Radcliffe Hospital); and Seattle, Washington (University of Washington Hospital, Swedish Hospital, and Providence Everett Hospital). Full acknowledgment of all those who contributed to the development of the INTERGROWTH-21^st Project protocol appears at http://www.intergrowth21.org.uk