Malnutrition may be defined as “a state of nutrition in which a deficiency, excess or imbalance of energy, protein, or other nutrients, including minerals and vitamins, causes measurable adverse effects on body function and clinical outcome.” In children, impairment of growth is clearly an adverse effect and is easily measurable by simple anthropometry. Because malnutrition can also be regarded as a continuum starting with a nutrient intake inadequate to meet physiological requirements, followed by metabolic and functional alterations, and ultimately by changes in body composition, classification based solely on commonly used growth criteria is not just arbitrary—it tells only part of the story.
Definitions in common usage for inpatients were developed by field workers in resource-poor countries attempting to determine the effect of food availability on local populations. How valid these are when considering the effect of nutritional status on disease outcomes in hospitalised children in developed countries is at least open to question. For example, where food is scarce, children between −1 and −2 SD deviations below the mean of weight for height (equivalent to Waterlow malnutrition grade 1) (1) may well be a vulnerable group (with increased risk of death from infection), but for many hospitalised children in developed countries with acute, short-term illness, this may indicate no more than having a thin physique or a temporary weight loss that will be made good within days of recovery and discharge home. For them to be classified as “at risk” (2) begs the question, “at risk for what?”
DEFINITIONS OF MALNUTRITION BASED ON GROWTH STATUS
Gomez (3) described nutritional status in children admitted to hospital in Mexico City in the 1950s and experiencing malnutrition from inadequate food availability. He divided them into 3 groups according to percentage weight for age (based on Boston growth standards) and showed that the most malnourished (weight for age <60%) were more likely to die, usually from respiratory or gastrointestinal infection. Gomez, therefore, established an important link between malnutrition, poor outcome, and the confounding factor of infection. Almost 20 years later, Waterlow discussed a new classification and definition of protein-energy malnutrition (PEM) (1). This was prompted by the Eighth Joint Expert Committee on Nutrition of FAO and WHO seeking a universal definition of PEM that would allow meaningful comparison of prevalence rates in different countries. Rather than just using percentage ideal weight for age, it was suggested that height or length measurements were important for giving an indication of the duration of malnutrition. Waterlow then subdivided deficit in percentage expected weight for height (thinness, indicative of acute malnutrition) and percentage expected height for age (stunting, indicative of chronic malnutrition). Grade 1 wasting (expected weight for height 90%–80% of the reference standard) highlighted vulnerable children but was not in itself considered to be an indication for intervention at a population level (4). McLaren and Read raised the objection that not all children of the same height should have the same weight because the relation between height and weight shows some variation with age (5). They devised a nomogram for diagnosing PEM and, in a worked example of >500 poor children from Beirut, showed that compared with Waterlow, their approach would triple the number of children requiring nutritional intervention. This graphically illustrates how minor changes in classification of malnutrition can lead to a large variation in estimated prevalence. The current WHO definitions of PEM categorise between −2 and −3 SD and <−3 SD, respectively, as moderate and severe acute malnutrition (weight for height) and moderate and severe chronic malnutrition (height for age).
PREVALENCE OF HOSPITAL MALNUTRITION IN DEVELOPED COUNTRIES
A number of studies during the last 30 years have attempted to establish the prevalence of “malnutrition” amongst hospitalised children in Europe and America using a variety of standards based on growth/anthropometry including Waterlow and WHO. Perhaps not surprisingly, given the use of different definitions and reference standards, different answers have been obtained, with ranges for chronic malnutrition varying, for example, from 9% to 47%. These studies share, however, the common conclusion that malnutrition in hospital settings is alarmingly common and often unrecognised. The need to evaluate this claim critically can be illustrated by reference to 1 recent study (6) in which the Waterlow criteria were used; if WHO criteria for malnutrition are substituted, the prevalence of malnutrition falls from 24% to around 6%. A recent Dutch national study (7) using WHO criteria found 11% of hospital admissions to show acute and 9% chronic malnutrition.
“SCREENING” FOR HOSPITAL MALNUTRITION
On the basis of the argument that paediatricians miss malnutrition because of an institutional incapacity to routinely weigh, measure, and chart all of the children, simple screening tools have been suggested; why the institution should be any better at using these is unclear. Alarmingly, 1 such tool has identified as many as 62% of children as being “at risk” for malnutrition. This is surprising given a median hospital stay of 2 days for those judged “low risk” and 3 days for “moderate and high risk” groups (8), suggesting that many were not severely ill. Implementing screening for arbitrarily defined malnutrition/risk when it is unknown whether screening tools in any way predict outcome or permit effective intervention may be premature. The dangers of such an approach include not only creating an unnecessary market for nutritional support products (1 study suggests prescribing sip feeds after screening but before medical and dietetic assessment) (8) but also that screening becomes a “tick box” indicator of imagined quality while in reality serving as a substitute for sound clinical assessment of individual children. If growth monitoring and nutritional history taking were performed routinely and linked to action plans for all children admitted to the hospital, then there would be no need for an alternative form of “screening.” The precise indications for nutritional support as well as the benefits of screening tools require scientific evaluation in specific groups of patients. Accurate growth measurements and plotting are certainly required because relying on clinical impression alone is highly inaccurate (9). Meanwhile, there should be individualised nutritional assessment for all children admitted to hospital, with interventions aimed at preventing or reversing growth deficits (Table 1) (10) and specific nutritional deficiencies.
1. Waterlow JC. Classification and definition of protein-calorie malnutrition. Br Med J 1972; 3:566–569.
2. Moy RJD, Smallman S, Booth IW. Malnutrition in a UK children's hospital. J Hum Nut Dietetics 1990; 3:93–100.
3. Gomez F, Galvan RR, Frenk S, et al
. Mortality in second and third degree malnutrition. J Trop Pediatr 1956; 2:77–83.
4. Waterlow JC. Some aspects of child nutrition as a public health problem. Br Med J 1974; 4:88–90.
5. McLaren DS, Read WWC. Classification of nutritional status in early childhood. Lancet 1972; 300:146–148.
6. Pawellek I, Dokoupil K, Koletzko B. Prevalence of malnutrition in paediatric hospital patients. Clin Nutr 2008; 27:72–76.
7. Joosten KF, Zwart H, Hop WC, et al
. National malnutrition screening days in hospitalised children in The Netherlands. Arch Dis Child 2010; 95:141–146.
8. Hulst JM, Zwart H, Cop WC, et al
. Dutch national survey to test the STRONGkids nutritional screening tool in hospitalized children. Clin Nutr 2010; 29:106–111.
9. Cross JH, Holden C, MacDonald A, et al
. Clinical examination compared with anthropometry in evaluating nutritional status. Arch Dis Child 1995; 72:60–61.
10. Braegger C, Decsi T, Dias JA, et al. Practical approach to paediatric enteral nutrition: a comment by the ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr