Background
IMCI provides clinical algorithms that guide health workers at primary health-care facilities on the assessment and management of infants and children presenting with illnesses, including assessment and classification of nutritional status. The algorithms reflect WHO guidelines on interventions to treat the most common and serious conditions causing mortality or serious morbidity among neonates and infants and children. An underlying principle of IMCI is that an assessment provides a classification and that classifications should translate into an appropriate care and treatment response. An assessment that does not result in a classification and relevant treatment would not therefore be included.
WHO classifications of nutrition conditions in children based on anthropometry are provided in the earlier section “Definitions and key terms”. Definitions of wasting are based on weight-for-length in children under 2 years of age or weight-for-height in older children, or an assessment of the mid-upper arm circumference. The presence of bilateral oedema is a further criterion for identifying children with severe acute malnutrition. WHO definitions for overweight and obesity are also based on both weight and length/height estimates, i.e. weight-for-length/height and body mass index (BMI) but not mid-upper arm circumference.
Children with severe wasting or severe acute malnutrition have significantly increased mortality risk, though appropriate treatment significantly reduces mortality (19, 20). Children with moderate wasting and/or stunting also have increased mortality risk, although less than for severe wasting or wasting and stunting combined (21).
In an individual patient meta-analysis of 53 607 participants and 1306 deaths, the risk of all-cause mortality was elevated among children with one, two or three anthropometric deficits. In comparison with children with no deficits, the mortality hazard ratios were 3.4 (95% confidence interval [CI]: 2.6 to 4.3) among children who were stunted and underweight but not wasted; 4.7 (95% CI: 3.1 to 7.1) in those who were wasted and underweight but not stunted; and 12.3 (95% CI: 7.7 to 19.6) in those who were stunted, wasted and underweight (21).
In children, moderate acute malnutrition (moderate wasting) may be transient and experienced by children in the course of an acute illness. Moderate wasting may also be due to chronic illnesses such as tuberculosis (TB) or HIV, or can also be the result of food insecurity and inadequate dietary intake. Management of moderate wasting should include a range of essential nutrition actions such as breastfeeding promotion and support, education and nutrition counselling for families, and other activities that identify and prevent the underlying causes of malnutrition, including nutrition insecurity. Interventions to improve food insecurity include the provision of conditional or non-conditional cash transfers and support to agriculture, such as crop diversification (22). Children with moderate wasting need nutrient-rich foods to meet their extra needs for full recovery (22). In the context of food insecurity, food supplements may have a role and food programmes have been established in many humanitarian crises or other settings of chronic food insufficiency.
Historically, anthropometric assessments and nutrition classifications in IMCI focused on identifying infants and children with undernutrition because of the associated health consequences. The classification of undernutrition included in IMCI focused on assessing weight-for-age because it is simpler to measure than the length or height of children. Children whose weight-for-age was less than the third percentile were classified as “very low weight”. However, this classification did not necessarily reflect the relevant risks of morbidity and mortality. For example, one child may have a very low weight-for-age because he or she is stunted or has a history of low birth weight. In contrast, another child who is taller may have the same weight-for-age and is wasted because he or she has suffered a recent acute or serious illness. Similar considerations are relevant for children and classifications for overweight and obesity.
Therefore, with respect to both wasting and overweight and obesity, it is important to relate a child’s weight-for-age to their length/height, in order to correctly interpret their nutritional status.
In the context of the double burden of malnutrition, it is therefore important that a public health approach, such as IMCI, correctly identifies children who are at risk of clinically important undernutrition (severe or moderate wasting, or severe acute malnutrition) and also children who are at risk of overweight and obesity.
Summary of the evidence
Three reviews of relevant information were completed to contextualize the environment in which recommendations may be implemented and to assist the guideline development group to make informed decisions. Details of the review teams and references are presented in Annex 1, Annex 2a and Annex 2b.
A meta-analysis of epidemiological anthropometric data of infants and children from different regions. This shows regional differences in rates of undernutrition and stunting in infants and children (
23).
A systematic review of the relationship between anthropometric measures in childhood and short- and long-term health outcomes (
24).
Note: Grading of Recommendations, Assessment, Development and Evaluation (GRADE) tables (
25) were not developed but comprehensive tables summarizing the evidence for associations between anthropometric measures in early childhood and anthropometric and health outcomes in adulthood are included in the report.
A survey of IMCI chart booklets and clinical algorithms used in different countries, reporting current uptake and adaptations of IMCI at national level (
26).
The first, an epidemiological analysis of surveys conducted in 63 low- and middle-income countries (1993–2012) (23) reported that between 0.4% and 18.6% of children (aged 6–59 months) would be classified as “very low weight”, i.e. weight-for-age more than 3 SD below the WHO child growth standards (2) (Bangladesh 14.1%, Chad 15.7%, India 18.6%, Niger 16.3%, Timor Leste 16%). The vast majority of these children are also stunted (low length/height-for-age). The proportion who were “very low weight” and also wasted varied by region. In Central and South American and Caribbean countries, 0.4–2% of stunted children were concurrently wasted, whereas in Africa and South Asia 0.5–11% were also wasted (see distributions by regions in Annex 1 (23)). The analyses highlight the regional differences in rates of undernutrition and stunting in infants and children, and the importance of correct classification of nutritional status for individual-based recommendations.
The second, a systematic review (24), analysed data from 52 studies (36 cohorts or longitudinal studies) that reported data on the association of anthropometric measures during infancy and childhood and the risk of developing obesity in later life (childhood, adolescence or adulthood). A meta-analysis using a random-effects model showed a significant inverse association of correlation coefficients. The pooled analysis of BMI showed a decreasing trend when tracked from early childhood (3–6 years) to mid childhood (9–11 years; r = 0.79, 95% CI: 0.71 to 0.87), adolescence (12–18 years; r = 0.66, 95% CI: 0.53 to 0.78), adulthood (19–40 years; r = 0.39, 95% CI: 0.29 to 0.50) and late adulthood (>40 years; r = 0.26, 95% CI: 0.18 to 0.34); mid childhood (9–11 years to adolescence (12–18 years; r = 0.81, 95% CI: 0.73 to 0.88), adulthood (19–40 years; r = 0.58, 95% CI: 0.50 to 0.66). BMI tracking showed similar trends in low- and middle-income countries and high-income countries. Thirteen studies showed that high BMI in childhood or adolescence was associated with a high risk of overweight or obesity in later life. The majority of the studies predicted a high risk of becoming overweight or obese with higher weight status during infancy and childhood.
Rapid weight gain in infancy and childhood was also strongly associated with the risk of overweight or obesity in adolescence or adulthood. Evidence from cohort studies showed that children who gained excess weight in “mid-child years” (age 4 years in four of the birth cohorts and age 8 years in one birth cohort) are at significant risk of obesity, diabetes and cardiovascular disease in adult life (27).
Eleven articles (including two consortia) from 13 cohorts discussed the associations of various anthropometric measures during infancy or childhood and cardiovascular morbidity and mortality in later life.
Four cohort studies showed that higher BMI during childhood was associated with cardiovascular disease morbidity and mortality during adult life. One cohort study (Helsinki Birth Cohort) showed that lower BMI during early childhood was associated with higher risk of cardiovascular disease during adult life. However, this finding was not consistent across all cohort studies. Two cohort studies, one each from Great Britain and the United States of America, did not find any association between childhood BMI and adult cardiovascular disease. The pooled results from three British cohorts found a significant relationship between being persistently overweight throughout childhood until adult life and coronary artery disease in later life. Pooled results from three historical British birth cohorts attributed minimal risk for ischaemic heart disease when childhood BMI was high. One cohort study (Helsinki) showed that lower weight and/or shorter length at 1 year of age increased the morbidity and mortality due to cardiovascular disease in adult life. Low weight-for-age at 1 year may have been due to low birth weight or being small for gestational age.
Findings regarding associations between childhood weight and height and systolic and diastolic blood pressure were not consistent. Among 29 283 subjects in 12 cohorts, 9 studies reported positive associations between BMI in childhood (2–12 years) and adolescence (12–18 years) and raised adult systolic and diastolic blood pressure measured between 19 and 50 years. One study, however, reported no association after adjusting for adult BMI.
In summary, there were strong correlations between high weight-for-length/height/BMI (and other markers of adiposity such as skinfold thickness) in early to mid-childhood and the prevalence of obesity, cardiovascular disease and diabetes in adolescents and adults (24).
The third review was a brief survey carried out by WHO in 2015 for the purpose of this guideline process (26), which found that anthropometric assessment of infants and children in the national IMCI chart booklet had been updated in 19 countries. In these versions, the combination of measurement most frequently included to assess nutritional status was weight-for-age + weight-for-length/height + mid-upper arm circumference, whereas in previous versions only assessments of weight-for-age were included. Clinical signs were used more frequently than anthropometric measurements to define acute malnutrition; “visible severe wasting” is still being used as a criterion in 39 countries despite no longer being recommended by WHO since 2013.
A separate growth-monitoring programme is carried out in 39 countries, mostly together with routine immunizations. Growth monitoring is most commonly performed in children below the age of 5 years. The anthropometric measurement used most frequently for growth monitoring was weight-for-age, then weight-for-length/height and length/height-for-age. Data on coverage of growth monitoring among children (up to 12 months of age) were available in 23 countries; high coverage of growth monitoring (80–100%) was reported in 12 countries (26).
Best practice statement
With respect to anthropometric assessment of infants and children presenting to primary health-care facilities, a statement on best practice was agreed.
Best practice statement: All infants and children aged less than 5 years presenting to primary health-care facilities should have both weight and length/height measured, in order to determine weight-for-length/height and to classify nutritional status according to WHO child growth standards (2).
Note: The measurement of mid-upper arm circumference both at health facilities and in the community can be used to identify children with severe or moderate acute malnutrition (19). However, mid-upper arm circumference cannot be used to determine overweight or obesity, as there are no validated cut-off values as yet. The best practice statement therefore only makes reference to weight and length/height.
Justification/remarks
The guideline development group considered that the statement did not constitute a recommendation that could be supported by comparative clinical studies. Instead, they considered it is self-evident that a child with acute undernutrition or who is overweight or obese must be identified, in order for appropriate interventions to be provided. No evidence was presented regarding the ability of health workers to effectively measure and interpret length/height and weight measurements to appropriately classify children. While growth monitoring is a common part of child health services, it is often challenging to do accurately (28, 29).
The group agreed, however, that they should make a formal statement that assessment of anthropometric status is important and provides the critical background against which several other important care and treatment decisions would be made. The group therefore decided that it was appropriate to formulate and present a best practice statement.
The group noted that using weight alone to inform clinical decisions leads to significant misclassification of nutritional status. One way to minimize this misclassification would be to standardize measurement of length/height in the context of IMCI. With this information, health workers would be able to provide appropriate nutrition counselling, e.g. information about food quality, and avoid inappropriate guidance such as advising parents whose children are underweight but of low length/height (therefore normal weight-for-length/height) to eat more.
An evidence base on health outcomes associated with reduced weight-for-length/height is available but there is no evidence on the impact of performing the assessment in routine conditions. Some members of the guideline development group felt that measuring length/height is not useful unless it is done properly. Also, it may not always be possible to measure length/height, and it could be an extra task carried out at the expense of some other health-worker activity.
Countries in the WHO Region of the Americas measure length/height routinely. Implementation was done with sensitization and advocacy for its importance, which has helped to put it into practice.
The arguments in favour of a recommendation were that if there were a recommendation to measure length/height routinely, then the strategy for implementation, including training, would be established.
Suggesting a “best practice” was considered to give flexibility to countries and reflect that measuring length/height may not always be possible.
Implementation considerations
Considerable initial training and feedback will be required to establish good measurement processes at health-care facilities.