Wasting in children

Author/s: Winnie Sambu & Katharine Hall
Date: February 2014


Wasting is also referred to as acute malnutrition, and is defined as low weight-for-height. Normally, a healthy child is expected to gain 2 – 3 kg of body weight every year. A child whose weight-for-height measurement is less than two standards deviation from the globally accepted reference cut-off point is considered to be wasted. Severe wasting occurs when the child’s weight-for-height measurement is less than three standard deviations from the globally accepted norm. 


Data Source Southern African Labour and Development Research Unit (2012) National Income Dynamics Study 2008, Wave 1 [Dataset]. Version 4.1. Cape Town, SALDRU, University of Cape Town (Producer), DataFirst, University of Cape Town (Distributor).
  1. Children are considered wasted when their weight-for-height z-score is less than -2 SD from the mean. Children are severely wasted when their z-score is less than -3 SD from the mean.
  2. This data refers to children aged under 5 years.

The effects of poor nutrition on children are far reaching. It is estimated that more than 200 million children under five years globally will not realise their full cognitive development due to poverty, lack of proper care, poor health and inadequate nutrition.1 Research suggests that poor nutrition affects the educational outcomes of children, adult working capacity and economic productivity.2 Under-nutrition in childhood could therefore lead to lower wages in adulthood, perpetuating intergenerational cycles of poverty and exacerbating poverty rates.

Globally, undernutrition contributes to more than a third of deaths in children under five.3 A local study of child deaths in audited hospitals indicated that 34% of children who died between 2005 and 2009 were severely malnourished and another 30% were underweight for their age.4 Early childhood is a critical period for growth and development, and nutritional deficits may be irreversible after the second year.5 The effects of early undernutrition are long-reaching, and are associated with life-threatening diseases such as diabetes, cardiovascular disease and hypertension in adult life.6

UNICEF distinguishes between the immediate, underlying and basic causes of malnutrition.7 Immediate causes of malnutrition include inadequate dietary intake and illness. This can lead to a potentially vicious cycle of illness and malnutrition, where malnutrition impairs children’s immunity leading to recurrent bouts of illness, which further undermine children’s nutritional status.8 Underlying causes include household food insecurity, inadequate maternal care, poor access to services and unhealthy living environments, which in turn are driven by the unequal distribution of resources in society.9 

Efforts to monitor malnutrition in South Africa are constrained by the shortage of regular and reliable anthropometric data (measures of height and weight, for example). Nationally representative surveys that have yielded usable data on the height and weight of children are the Project for Statistics on Living Standards and Development (PSLSD) of 1993, the Demographic and Health Survey of 1998, the National Food Consumption Survey of 2005 and the National Income Dynamics Study (NIDS) of 2008.

It is notoriously difficult to collect anthropometric data of good quality. Statistics South Africa’s Living Conditions Survey of 2008/09 collected anthropometric data from a large sample but did not publish it because the quality was too poor. Subsequent iterations of the NIDS panel survey have collected anthropometric data, but although changes in children’s nutritional status over time are plausible,10 the representivity of the sample diminishes after the first wave. The analyses presented here are therefore based on the most recent reliable and nationally representative data: NIDS 2008. A more recent survey, the South African National Health and Nutrition Examination Survey,11 was undertaken in 2012, and may provide more up-to-date data for analysis of child anthropometry. The data have not yet been made available.

Unless otherwise specified, the results are based on analyses of NIDS (2008) and the PSLSD (1993) by Winnie Sambu of the Children’s Institute, UCT. In both cases, the malnutrition rates have been derived based on the World Health Organisation’s Child Growth Standards.12

Wasting is caused by infection and inadequate nutrition. It can change rapidly depending on the availability of food and the presence of illness, and is therefore a measure of acute (rather than chronic) malnutrition.13  

In 2008, 4.8% of children under five years were wasted, and 2% severely wasted. Wasting rates were highest in urban informal areas (8%) followed by the former homelands (4%). No statistically significant differences in wasting were found amongst male and female children. The incidence of wasting and severe wasting has declined since 1993, when the rates were 9% and 4% respectively.

As part of NIDS (2008), anthropometric measurements were recorded. Two height and weight measurements were collected from each child, and a third one if the one and two sets of measurements were more than one centimetre or one kilogram apart respectively. An average of the first two measurements was in each case taken for the purposes of Z scores derivation while the third measure was used for Z scores derivation if the first two were more than centimetre or one kilogram apart in the height and weight measurements respectively. The weights and heights collected during the study were converted to Z scores based on the 2006 WHO international child growth standards for children aged up to 5 years and in the of children aged more than 5 years, the WHO growth standards for school going and adolescent children was used.14

NIDS attempts to follow individuals and changes that occur in their well-being over a period of time. Wave 1 data collection began in February 2008, and involved 7,305 households and 28, 255 individuals. The study used a two-stage cluster sampling approach, where in the first stage; primary sampling units were selected from Stats SA’s master sample. During the survey, data collected included household demographics, income and expenditure patterns, living conditions, anthropometric measurements among other indicators. While it is a nationally representative survey, further disaggregation is limited due to the small sample size used.