Article Text

Download PDFPDF

Conceptual issues related to the construction of prescriptive standards for the evaluation of postnatal growth of preterm infants
  1. J Villar1,
  2. H E Knight1,
  3. M de Onis2,
  4. E Bertino3,
  5. G Gilli3,
  6. A T Papageorghiou1,
  7. L Cheikh Ismail1,
  8. F C Barros4,
  9. Z A Bhutta5
  10. for the International Fetal and Newborn Growth Consortium (INTERGROWTH-21st)
  1. 1Nuffield Department of Obstetrics and Gynaecology, and Oxford Maternal and Perinatal Health Institute, Green Templeton College, University of Oxford, Oxford, UK
  2. 2Department of Nutrition for Health and Development, World Health Organization, Geneva, Switzerland
  3. 3Dipartimento di Scienze Pediatriche e dell' Adolescenza, Cattedra di Neonatologia, Università degli Studi di Torino, Torino, Italy
  4. 4Programa de Pós-Graduação em Saúde e Comportamento, Universidade Católica de Pelotas, Pelotas, RS, Brazil
  5. 5Division of Women and Child Health, The Aga Khan University, Karachi, Pakistan
  1. Correspondence to Dr J Villar, Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Women's Centre, Level 3, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK; jose.villar{at}


Monitoring and interpreting the growth of preterm infants is a major clinical task for neonatologists. The effectiveness of this process depends upon the robustness of the standard selected. Concerns have been raised regarding the nature of the charts currently being used, as well as their appropriateness for present-day neonatal care. To overcome these problems, there is a need for new prescriptive standards based on a population of preterm infants without evidence of impaired fetal growth and born to low-risk women followed up since early pregnancy for precise gestational age dating. Preterm infants contributing to the new standards should be free of congenital malformations and major clinical conditions associated with impaired postnatal growth. These infants should receive standardised, evidence-based clinical care and should follow current feeding recommendations based on exclusive/predominant breastfeeding. This strategy should provide a population that is conceptually as close as possible to the prescriptive approach used for the construction of the WHO infant and child growth standards. New international standards constructed in this way should contribute to the evidence-based care of these preterm infants.

View Full Text

Statistics from


Preterm birth is a leading public health problem worldwide; it is responsible for up to 30% of neonatal deaths, and indirectly responsible for a further 20%.1 Its rates are increasing worldwide.2 Although improved perinatal care has led to major increases in survival even at very low gestational ages, long-term complication rates have unfortunately remained stable.3

Concerns have been raised regarding the strategies to monitor the postnatal growth of preterm infants, specifically in relation to the charts used.4 Growth charts are essential tools in monitoring the pattern of postnatal growth; they are used at the individual and community level to ensure timely and adequate nutritional intervention, referral, and treatment of individuals and populations. This is particularly important given the association of postnatal growth restriction in infants born preterm with lower cognitive and academic scores at school.5 6 Although there are several anthropometric measurements for which growth charts are currently available, weight gain is often the measure most used in very young infants.

Five strategies exist for monitoring the postnatal growth of preterm infants: (a) fetal growth curves based on intrauterine ultrasound measurements; (b) birth weight for gestational age charts; (c) postnatal longitudinal growth charts for preterm infants; (d) prescriptive growth standards for infants born at term (37–41 weeks gestational age); and (e) a combination of (b) and (d). We will briefly discuss these strategies in turn, none of which are without their problems, before describing a strategy that could potentially overcome these challenges.

Current tools for monitoring preterm postnatal growth

Fetal weight estimation curves based on ultrasound measurements

Estimated fetal weight charts, or ‘ideal weight’ charts, are based on ultrasound measurements of fetal anthropometry at various gestational ages. They may not be appropriate because estimated fetal weight is, by definition, an estimation of actual weight and large measurement errors are known to exist.7 8 Furthermore, using this type of intrauterine chart tends to show that by the time preterm infants reach ‘term’ they are ‘intrauterine growth restricted’, even if catch-up growth has occurred.

Interestingly, improvements in the survival of very preterm infants have been reported without reaching the suggested ‘fetal growth’ rates postnatally. In addition, these ‘intrauterine’ growth charts do not consider the physiological weight loss that occurs in the neonatal period (up to 8% of the birth weight in the first week of life)9 particularly among term babies below the 25th percentile of birth weight10 and preterm infants weighing <1500 g at birth.11

Birth weight for gestational age charts

The second method used to monitor postnatal growth is the birth weight for gestational age ‘neonatal chart’. Several of these charts are available, ranging from single hospital samples to country-based populations. Most of these cannot be considered prescriptive standards. Data used to construct them are cross-sectional and do not represent the actual growth of preterm infants that occurs during extrauterine life as they are based on anthropometric measures taken once (at birth) in infants born at different gestational ages. As such, they should be considered ‘size-at-birth charts’ and should not be used for monitoring growth; they can neither be extrapolated to the growth of healthy fetuses that remained in utero12 nor to the postnatal growth of preterm infants.

Postnatal longitudinal growth charts for preterm infants

Considerable efforts have been made to document the actual postnatal growth of very low birthweight babies. Despite this very important work, many cannot be considered a contemporary standard for clinical use.13,,15 Firstly, most of them concentrate on very low birthweight infants classified by birthweight intervals, not considering their gestational age, and therefore including intrauterine growth restriction infants. Secondly, they seldom have early ultrasound dating or reliable last menstrual period date.16 Thirdly, cases with maternal pathologies that could affect fetal growth and lead to newborn complications are included. Fourthly, they do not specify feeding practices. Finally, most charts have relatively short follow-up periods. Nonetheless, the experience gained by these previous studies is of considerable value for the present conceptualisation of the strategies for constructing appropriate prescriptive standards.

Prescriptive growth standards for infants born at term

After reaching postmenstrual term, preterm infants can be monitored using the new prescriptive postnatal standards for term infants, which have now been adopted by over 100 countries.10 17 However, when growth parameters of preterm infants are plotted, these measures often fall far below the lowest centiles or Z scores, as the postnatal growth of preterm is different from that of term infants. This is therefore a strong argument for the need to construct a prescriptive growth standard specifically for preterm babies.

Combining birth weight for gestational age charts with prescriptive growth standards for term babies

A combination of a birth weight for gestational age reference and a postnatal growth reference has been suggested.18 This combination still has the same fundamental problems listed above for each of its components except a larger sample size for the very early gestational age ranges. In addition, the postnatal chart for term infants used by Fenton18 is now being replaced by the new WHO standards.17 Finally, we consider that the ponderal index is an important addition to the nutritional assessment of preterms infants.19 20 A comprehensive review has been recently published, complementing this discussion.16

Constructing new preterm growth standards

We now present the conceptual issues that will be important in guiding the construction of a new ‘prescriptive’ preterm growth standard.

Concentrate on preterm deliveries

The first recommendation is to concentrate only on preterm deliveries (below 37 weeks of gestational age) and not use the proxy of low birth weight. Low birth weight represents a heterogeneous group of conditions with different intrauterine, postnatal growth and morbidity and mortality characteristics.21 22 Selecting newborns by birth weight will lead to an over-representation of small for gestational age or intrauterine growth restriction babies, which should be studied separately.33

Prescriptive, rather than descriptive standards

Conceptually speaking, standards should be prescriptive.17 They should not describe how preterm infants have grown postnatally at a particular time and place, but rather how they should grow under optimal postnatal conditions.17 24 We propose to extend this concept to develop new standards for fetal, newborn and postnatal preterm growth and relate them to morbidity and mortality indicators as described in the INTERGROWTH-21st study protocol.25 This is a major conceptual jump, aiming to improve upon the shortcomings of existing growth charts for infants born preterm.

To achieve this aim, standards must be based on a population that has the lowest risk for factors known to affect prenatal and/or postnatal growth, or length of gestation26 17 with pregnancy and neonatal outcomes according to this low risk. This principle does not attempt to select a ‘perfect’ population, as it is clear that even late preterms will have some level of complications or developmental delay, but rather one that has the best potential for growth.

Inclusion criteria

Which preterms should therefore contribute to the new standard? These newborns should be born to women at the lowest risk for fetal growth restriction (box 1). This low-risk pregnancy background is important as preterm infants born to mothers without pregnancy complications are at the lowest perinatal morbidity and mortality risk as compared with other preterm subgroups.21

Box 1 Inclusion criteria for the mothers of the preterm infants

  1. (A) Aged ≥18 and ≤35 years

  2. (B) BMI ≥18.5 and <30 kg/m2

  3. (C) Height > 153 cm

  4. (D) Singleton pregnancy

  5. (E) A known last menstrual period with regular cycles (defined as 28 ± 4 days) without hormonal contraceptive use or breastfeeding in the 2 months before pregnancy

  6. (F) Natural conception

  7. (G) No relevant past medical history, with no need for long-term medication (including fertility treatment and over-the-counter medicines, but excluding routine iron, folate, calcium, iodine or multivitamin supplements)

  8. (H) No evidence of socioeconomic constraints likely to impede fetal growth identified using local definitions of social risk

  9. (I) No use of tobacco or recreational drugs such as cannabis in the 3 months before or after becoming pregnant

  10. (J) No heavy alcohol use (defined as >5 units (50 ml pure alcohol) per week) since becoming pregnant

  11. (K) No more than one miscarriage in the two previous consecutive pregnancies

  12. (L) No previous baby delivered preterm (<37 weeks) or with a birth weight <2500 or >4500 g

  13. (M) No previous neonatal or fetal death, previous baby with any congenital malformations, and no evidence in present pregnancy of congenital disease or fetal anomaly

  14. (N) No previous pregnancy affected by pre-eclampsia/eclampsia, HELLP syndrome or a related pregnancy-associated condition

  15. (O) No clinically significant atypical red cell alloantibodies

  16. (P) Negative urinalysis

  17. (Q) Systolic blood pressure <140 mm Hg and diastolic blood pressure <90 mm Hg

  18. (R) No diagnosis or treatment for anaemia during this pregnancy (Hb levels will be monitored throughout pregnancy)

  19. (R) No clinical evidence of any other sexually transmitted diseases, including syphilis and clinical trichomoniasis

  20. (T) Not in an occupation with risk of exposure to chemicals or toxic substances, or very physically demanding activity to be evaluated by local standards. Also women should not be conducting vigorous or contact sports, as well as scuba diving or similar activities

BMI, body mass index.

Secondly, although it is clear that only very few extreme preterms are without some type of severe clinical or laboratory complications when compared, for example, to preterms >32 weeks, all preterms (including those <26 weeks) should be considered for inclusion as long as they meet the morbidity criteria presented below. Many of the complications seen in preterms are not pathological as such, but rather the product of the physiological limitations that are to be expected due to the functional immaturity of their organs or systems for the needs of extra-uterine life.

Typical examples of immaturity-related conditions are respiratory distress syndrome (RDS) and jaundice. This concept of ‘expected normality’ for a given age is always implicit when making evaluations of this nature. For example, the concept of ‘developmental milestones’, is used to evaluate an infant's neurological development.27 Infants are evaluated for skills that should be present, according to the level of development expected at a given age range: for example, by 6 months of age an infant would not be expected to stand alone, even if that is normal later in childhood. Applying this same concept to very preterm newborns, even a relatively ‘healthy’ preterm newborn cannot be expected to be without some type of morbidity.

Threshold values for ‘normality’ are somewhat arbitrary and can be different for each pathology with a continual decline in their prevalence rather than a break point at a given gestational age. For example, the incidence of intraventricular haemorrhage and necrotising enterocolitis (NEC) gradually reduces and by 30–32 weeks is a rare condition, whereas the risk of RDS remains significant until 34 weeks.28 Apnoea of prematurity, on the other hand, is a developmental disorder that generally disappears by about 36–40 weeks gestational age.28

Further work is needed to properly define the gestational age at which each moderate condition can no longer be considered ‘normal’. One possibility is to extend the developmental milestones framework to cover preterm growth and development. For example, by 34 weeks postmenstrual age an uncomplicated infant should not require assisted ventilation due to RDS.29

Table 1 presents a classification according to the severity of complications related to the levels of organ maturation. There is limited published evidence that any of the conditions presented in the third column are associated with postnatal growth restriction. Rather, these conditions are primarily associated with a lack of complete organ maturation including physiological functions required for extra-uterine life. The fourth column indicates the pathological conditions that are known to affect postnatal growth, and should thus be excluded.

Table 1

Criteria for evaluating preterm newborns according to complications associated with postnatal growth restriction

Exclusion criteria

Exclusion criteria for infant candidates should include diseases or conditions that are mostly unrelated to the level of maturation, such as chromosomal disorders, hydrops, major congenital malformations, inborn errors of metabolism or evidence of fetal growth restriction from serial fetal ultrasound growth (eg, ultrasound measurements of abdominal circumference below the fifth percentile). Subjects with severe complications such as sepsis, NEC stages IIB and III, broncho-pulmonary dysplasia30 or death should also be excluded.

Preterm newborns should not contribute to the standards if they require any surgery. It has been suggested that during acute metabolic stress, neonates divert protein and energy from growth to tissue repair. This may explain the impairment of growth observed in infants after surgery.31 Whether other surgery, for example ligation of patent ductus arteriosus, affects postnatal growth remains under discussion in the literature.32 However, for the purposes of constructing prescriptive preterm standards, any newborn who undergoes surgery should be excluded from the standards.

Considerable discussion exists in the literature about the possible effects of other preterm-related conditions on postnatal growth (eg, intraventricular haemorrhage,33,,35 cystic leukomalacia,36 RDS,37 gastroesophageal reflux,38 apnoea of prematurity,39 NEC35 40,,42). We have decided that there is sufficient clinical evidence to suggest that severe forms of these conditions could influence postnatal growth and they should therefore be excluded when constructing prescriptive growth standards from the population to be included in the standards (table 1). In addition, as it is impractical to list all the possible complications that could arise in the cohort, the WHO/Multicentre Growth Reference Study strategy of appointing an external clinical review group, should be followed to evaluate individual cases with conditions that were not classified in advance.

Standardised nutritional strategies

There is marked heterogeneity in the nutritional management of preterm newborns among neonatal units.43 44 For example, recommended caloric intakes can range from 110 to 135 kcal/kg of body weight. It is expected that such variation will be present in local practice within the recommended ranges of doses or intakes. This should ensure that the resulting percentiles incorporate both the clinical variability of the presently used guidelines as well as the normal variation in growth.

Considerable efforts should therefore be made to ensure that the infants in the cohort adhere to internationally validated recommendations for parenteral45 and enteral nutrition.46 This includes exclusive or predominant breastfeeding and potential human milk fortifiers in extreme preterm infants. The international recommendations for feeding preterms were made by reaching consensus based on the best available science, using outcome-based evidence. Thus, despite some conceptual or practical limitations, these recommendations are at the present time, the most logical for the proposed study.

Standardised neonatal care strategies

All preterm infants should receive standardised evidence-based medical care across neonatal units and the follow-up clinics of the study. This can be achieved by coordination among the directors of the neonatal units following a manual of operation for clinical practice developed specially by the study's Neonatal Group (

Outcome information

For all preterm infants studied, detailed information should be collected on fetal growth before 14 weeks of pregnancy, pregnancy complications, neonatal events and postnatal growth and morbidity up to 8 months of life. We are requesting additional funding to complete the follow-up to 2 years of postnatal life. These data should allow, albeit with a relatively short follow-up period, to evaluate how their growth trajectories related to their clinical and neurological status, reinforcing the concept of ‘prescriptive’ charts.

Multi-ethnic origin

The inclusion of populations from different geographical regions is supported by the recent WHO standards that pooled six ethnic groups around the world47 and by the limited racial differences observed in the US 1995–2002 newborn population (23 g 95% CI: 8 to 38 g) between the birth weights of African American and white infants when studying the relationship between low fetal optimal weight and neonatal morbidity and mortality.48 These authors concluded that “because such a small difference may not be clinically relevant and possibly also confounded by uncontrolled factors, (we) do not believe that this study supports separate fetal growth standards by race, educational status, parity, smoking and maternal age”.48 The use of the so-called fetal or neonatal ‘customised growth standards’ by ethnic groups, parity or maternal height, are not only of minimal clinical or epidemiological justification49 but also of limited practical implementation on today's multi-ethnic perinatal care services.50

All these criteria (box 2) should together provide a population that is conceptually as close as possible to the prescriptive approach used for the construction of the new WHO child growth standards.17 Once the data has been collected, recommended statistical methodology should be applied to construct the new preterm growth standards.51


Multiethnic, population-based, prospective data collected under recent medical care

Healthy, well-nourished maternal population

Early evaluation of gestational age confirmed by ultrasound examination before 14 weeks

Inclusion only of preterm deliveries (not using low birth weight as a proxy)

Prospective ultrasound measures of fetal growth to exclude fetuses with evidence of impaired fetal growth

Preterm infants included only if they do not have major neonatal complications, neonatal surgery, congenital malformations or death in the complete follow-up period

Standardisation of feeding practices and newborn care among study centres

Standardisation of anthropometric measurements including use of the same equipment and techniques

Frequency of anthropometric measurements close during periods of fast growth (eg, every 2 weeks during the first 2 months)

Follow-up period during infancy to allow interface with WHO child growth standards

Adequate sample size for each range of gestational ages to allow presentation by Z scores and centiles


Preterm infants are at high risk of long-term health and developmental sequelae, including also late preterm infants born in settings with advanced perinatal care and adequate postnatal circumstances.52 53 Considering its complex aetiology and phenotypic characteristics it is unlikely that a single preventive or treatment modality will reverse this pattern in the near future.2 Monitoring and interpreting physical growth and development of preterm infants represents a major clinical task and it depends upon the robustness of the standard selected. We hope that international standards will contribute to the evidence-based care of these babies.


View Abstract


  • Funding Bill and Melinda Gates Foundation.

    Mercedes de Onis is a staff member of the World Health Organization. The author alone is responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy or views of the World Health Organization.

    Aris Papageorghiou is supported by the Oxford Partnership Comprehensive Biomedical Research Centre with funding from the Department of Health NIHR Biomedical Research Centres funding scheme.

  • Competing interests None.

  • Ethics approval Ethical approval for this study was granted by the Oxfordshire Ethics Committee C in December 2008.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.