New population cut-offs for classifying child malnutrition levels
Public Health Nutrition Editorial Highlight: ‘Prevalence thresholds for wasting, overweight and stunting in children under 5 years’ Authors: Mercedes de Onis, Elaine Borghi , Mary Arimond, Patrick Webb, Trevor Croft, Kuntal Saha, Luz Maria De-Regil, Faith Thuita, Rebecca Heidkamp, Julia Krasevec , Chika Hayashi and Rafael Flores-Ayala discuss their research below.
As of 2018, WHO and UNICEF have started using new prevalence thresholds for child stunting, wasting, and overweight in official reporting. In the current issue of Public Health Nutrition, we report on the development of the new thresholds.
Since the 1990s, prevalence ranges for undernutrition have been used by the international nutrition community for mapping countries according to levels of severity of malnutrition, as well as to identify priority countries for action. Countries also rely on these ranges for monitoring purposes and to trigger actions aimed at achieving national targets.
In 1993, prevalence ranges for stunting were determined by categorizing low- and middle-income countries by observed quartiles: “low” (<20%); “medium” (20-29%); “high” (30-39%); and “very high” (≥40%). For wasting, prevalence ranges were derived based on the association to under-5 mortality rates in 42 refugee camps: “acceptable” (<5%); “poor” (5-9%); “serious” (10-14%); and “critical” (≥15%).
Since then, important developments occurred: 1) with ongoing global reduction of stunting, half of the countries have approached a level below 20%, designated as “low”; 2) new WHO Child Growth Standards were released, affecting survey estimates; and 3) the need for a similar classification for overweight has become evident given rising trends and contributions to disease burdens. These changes prompted a re-examination of existing prevalence levels and the development of new ones for overweight. The WHO-UNICEF Technical Expert Advisory Group on Nutrition Monitoring (TEAM) was commissioned with this task including harmonization of terminology to label different categories.
Three approaches were considered: 1) a “descriptive approach” similar to the method used in the 1990s for stunting, where prevalence levels are established based on descriptive analyses grouping the nationally representative anthropometric estimates into categories corresponding to observed quartiles; 2) a “functional approach” similar to the method used in the 1990s for wasting, where prevalence levels are based on associations with mortality rates or other health outcomes ; and 3) a “novel approach” setting prevalence levels in relation to standard deviations (SD) of the new WHO Child Growth Standards.
The “novel approach” was chosen and the definition of ‘normal’ (i.e., within two SD from the WHO standards median) was used to define the first threshold “very low” (including 2.3% of the area under the normalized distribution). Multipliers of “very low” (rounded to 2.5%) set the basis to establish subsequent thresholds. Labels were harmonized across indicators as “very low”, “low”, “medium”, “high”, and “very high. To evaluate the implications of these new thresholds, nationally representative surveys from 134 countries were used to assess country groupings.
A key use of the prevalence thresholds is to trigger action aimed at achieving “low” or “very low” levels. The authors acknowledge this will require identification of recommended actions to be taken at each level for each nutrition conditions. To date only concrete programmatic actions for “high” and “very high” levels of wasting have been recommended in the context of the management of nutrition in major emergencies. A revision of these actions, based on new programmatic evidence, and development of recommendations for “low”, “medium”, “high”, and “very high” wasting, overweight, and stunting, is an immediate future need.
Read the full Open Access article here ‘Prevalence thresholds for wasting, overweight and stunting in children under 5 years’