Undue focus on body size

Being severely small (short or thin) alone is unimportant. What matters is the associated risk of severe adverse outcome (eg, death/disease/developmental delay). Simplicity makes anthropometry useful for epidemiological and clinical case-definitions. Often forgotten is that it is an indirect measure, akin to a proxy measure, an indicator of an underlying problem, analogous to spots in measles or fever in malaria. Better measures of nutritional status such as body composition are available,5 but are costly and difficult to do in most field settings. Anthropometry is valuable but has imperfect:

• Sensitivity: current global metrics focus on population extremes, with terms in table 1 often used interchangeably. Not differentiating state from process identifies most children with malnutrition but misses those with ‘normal’ anthropometry who are deteriorating and are smaller than they would be in a resource-rich environment and therefore at risk of adverse outcomes (false negative). Thus, true malnutrition-associated burden of disease is underestimated. Neglecting process also risks undue focus on treatment rather than prevention since actions may not occur until a child deteriorates below threshold values.

• Specificity: some children are constitutionally small but have no functional/clinical impairment or increased risk (false positive). In others, the underlying problem is not lack of food but another issue (eg, repeated infection, chronic inflammation and disability). Feeding may temporarily increase weight, but relapse follows if the underlying cause is not addressed.

Nutrition ‘tribalism’

Different anthropometric deficits are commonly managed by different communities. UN/NGO actors are especially prone to division:

• Stunting, often equated with ‘chronic malnutrition’, is high priority for development actors.

• Wasting, often equated with ‘acute malnutrition’, is high priority for humanitarian actors.

• Underweight is widely measured in Growth Monitoring Programmes but not elsewhere, hence impeding interprogramme referrals.

• Micronutrient malnutrition is often neglected in the above.

This is unfortunate, as actions on common risk factors and common determinants are too easily missed or diluted.

Inadequate consideration of context

Different individuals and populations may have the same anthropometry but varying levels of risk. Context details are important, yet are often overlooked, for example,

• Humanitarian emergencies can lead to sudden deterioration in weight so that even a mild/moderate deficit has high associated risk.

• Suboptimal but stable contexts can be associated with populations/individual adaptation and may not have correspondingly high morbidity/mortality.

Focus on severe adverse outcomes

An important clinical maxim is to ‘treat the patient, not the numbers’. The ultimate aim of any intervention is to ensure children not only survive but thrive. Severe in our framework refers to clinically important severe adverse outcomes: mortality, morbidity, impaired development. To capture true programme impact, these should be measured wherever possible (since some interventions might improve anthropometry but not mortality;6 others improve nutrition but not anthropometry).7 The framework also highlights underlying clinical causes because acute/chronic illness and anthropometric deficits are intimately related: deficits are associated with increased risk of and impaired response to illnesses; illness can exacerbate anthropometric deficit.

Inclusivity (including of long-term consequences)

Different manifestations of malnutrition are more-common or less-common/important in different settings but details are mainly relevant to technical audiences. While widely understood to mean undernutrition, we acknowledge that malnutrition is also used to describe overweight. Severe overweight is widely recognised as obesity so our using ‘severe malnutrition’ does not risk misunderstanding. It does create opportunities to highlight long term consequences of undernutrition: speaking about mal rather than undernutrition is a reminder that early-life undernutrition increases susceptibility to the harmful effects of adult overweight, especially elevated non-communicable non-communicable disease.8

Increased linkages and synergies between different nutrition-related programmes

Working towards common mortality/morbidity goals could facilitate interprogramme linkages and more efficient resource utilisation.

‘Severe malnutrition’ also provides bridging language supporting recent research on relationships between different malnutrition manifestations:

• Wasting predisposes to stunting9—‘chronic’ terminology under-recognises the role of acute insults.

• Wasting is common in children with low birth weight and disability10 and many wasted children are stunted—‘acute’ terminology is usually erroneous.

• Children with concurrent wasting/stunting are at particularly high risk of death11—programmes should be integrated.

Since all forms of malnutrition are fundamentally social problems of poverty and inequity, our terminology also emphasises the importance of tackling these common determinants.12

Clarity and familiarity

Both ‘severe’ and ‘malnutrition’ are widely understood terms and are easily translatable into most languages worldwide. ‘Severe malnutrition’ also has historical track-record, having been used alongside older case definitions with different growth references and classifications predating table 1. Children admitted to today’s malnutrition programmes differ to those in years past, but common characteristics are increased mortality/health risks compared with those without anthropometric deficit.

Keeping technical arguments internal

As in any field, some issues are novel and/or contested for example.

• Use of weight-for-age to identify concurrent wasting/stunting.11

• Mid-upper arm circumference versus weight-for-height to identify children for therapeutic feeding.13

• Recent focus on ‘wasting’ terminology neglects kwashiorkor.14

An overarching term provides stability and prevents important technical arguments giving the impression to those outside the sector that experts do not agree on other core issues.

Flexibility for programmes to use locally appropriate admission criteria

Guided by local mortality/morbidity data, individual countries could be freer to set context-specific programme admission criteria for ‘severe malnutrition’.

• In some settings, even mild/moderate anthropometric deficits are associated with excess mortality: integrated programmes treating both severe and moderate wasting as a continuum are hence important.15

• In other settings, mortality associated with anthropometric deficit is lower - costs of therapeutic/supplementary foods may not justify their benefits.16

For most technical purposes, quantifying and monitoring global trends and comparisons, current metrics focusing on table 1 terminology remain valid. However, discussions are needed on measuring risk; moving beyond the acute/chronic malnutrition dichotomy; feasibility of improved metrics17 and calculating a single global estimate of the burden of severe malnutrition.