Assessment of Nutritional Status in Emergency-Affected Populations - Adolescents (UNSSCN, 2000, 24 p.) |

Anthropometry is the measurement of certain parameters of the
human body. It is frequently used to assess nutritional status in young children
and adults ^{12, 20, 21}. Anthropometry has also been used to study the
growth and development of school-aged children and adolescents. Only recently
has an attempt been made to use anthropometric methods to assess acute
undernutrition in adolescents ^{12}.

Use of anthropometry requires two essential items: an anthropometric indicator and a cut-off point. The indicator, often called an anthropometric index, is a measurement or a combination of measurements made in the field, such as weight and height, or the combination of measurements with additional data, such as age. Different indices reflect different components of nutritional status. The index weight-for-height indicates thinness, and because acutely undernourished persons generally lose body weight but not height, weight-for-height decreases with acute undernutrition. On the other hand, young children with chronic undernutrition may not be thinner than normal children, but may have retarded growth in height. Chronic undernutrition may not be severe enough to cause weight loss, but does interfere with normal linear growth. As a result, height-for-age is decreased, and children become stunted. Weight-for-age reflects both acute and chronic undernutrition because both thin children and stunted children are underweight. In many emergency-affected populations, acute undernutrition may be superimposed over a high level of background chronic undernutrition. As a result, both thinness and stunting may be common.

Mid-upper arm circumference (MUAC) is an indicator of the amount of fat and muscle in the upper arm. Skinfold thickness measurements, taken at various places on the body, provide an estimate of the thickness of subcutaneous fat. Acutely undernourished persons metabolise fat and muscle to compensate for decreased nutrient intake, resulting in a decline in skinfold thickness and MUAC.

Anthropometry can be used to evaluate either individuals or
populations. To identify those in need of nutritional rehabilitation, a cut-off
point is established below which persons are offered nutritional therapy. Young
children 6-59 months of age with severe acute undernutrition are usually treated
in inpatient therapeutic feeding centres. The recommended admission criteria are
weight-for-height <-3 z-scores or <70% of median (see Box 1 for
explanation of cut-off points) or the presence of oedema ^{14, 22}.
Anthropometric measurements are taken frequently during nutritional therapy and,
in combination with clinical observations, are used to determine when children
can be discharged. The application of universal cut-off points has the dual
advantage of allowing comparisons of the level of undernutrition between
populations and also helping to prevent bias on the part of feeding-centre staff
when performing initial assessment or follow-up of patients.

Anthropometry is also used to determine the prevalence of undernutrition in a population. Anthropometric measurements for each child selected as part of a representative sample are compared to a reference population to determine each child's nutritional status. The proportion of sampled children who are undernourished provides an estimate of the prevalence of undernutrition in the entire population of children. Such surveys are most commonly performed in children 6-59 months of age, but can be undertaken in any population subgroup, such as older adults, adolescents, or pregnant women. Estimates of the prevalence of undernutrition, along with other data on food and health, are used to plan programmes of food aid and nutritional therapy, or to evaluate the effect of such programs.

Anthropometry, however, cannot provide the complete picture of
the nutrition and food situation needed for problem solving and programme
planning. Anthropometry can provide an estimate of the prevalence of
undernutrition, but evaluations of food security, food distribution, nutrient
content, morbidity and mortality, and other elements are needed to understand
the causal factors resulting in undernutrition ^{23}. Nonetheless,
anthropometric measurements are relatively easy to obtain in the field and
anthropometric surveys can often be carried out in displaced populations, even
during the acute phase of a humanitarian emergency.

For each anthropometric index, a specific level must be determined as the cut-off point which distinguishes the normal nutritional state from undernourished. Cut-off points are also used to distinguish different levels of undernutrition. Cut-off points for anthropometric indices can be determined statistically. For example, the most commonly used index of acute undernutrition in young children is weight-for-height. Each measured child is compared to the reference population to determine how far that child is from the average child in the reference population. This discrepancy between an individual child and the reference can be expressed in a number of ways. One frequently used method when assessing individual children is the percent of median. A weight of a measured child is compared to the median weight of all children in the reference population of the same sex and height. We will use as an example a 17 month-old girl who weighs 8 kg and is 80 cm long. The median weight of girls in the reference population who are 80 cm long is 10.6 kg (note that age is not necessary). Therefore, the measured child has a weight-for-height which is 75% of median. An alternate method, now somewhat outdated, is to express weight-for-height in centiles. That is, upon looking up in a centile table, we find that our 8 kg, 80 cm girl weighs less than 84% of girls in the reference population who are 80 cm long. Therefore, this child falls on the 16th centile of weight-for-height. In population surveys, the preferred method for expressing the comparison between a specific child and the reference is by using z-scores. The distribution of weights for all reference children of the same sex and height is described by the median and standard deviation, given in kg. One standard deviation is one z-score. Using the example girl referred to above, the standard deviation of the weight for reference girls who are 80 cm long is approximately 0.88 kg. The median weight of these reference girls is 10.6 kg. Because the weight of our example child is 2.6 kg lower, she falls 2.95 standard deviations below the median. Therefore, her weight-for-height z-score is -2.95. The z-score cut-offs for moderate and severe undernutrition (<-2 and <-3, respectively) were determined from the distribution of weight-for-height values in a sample of American children. It must be remembered that reference populations are used only to develop statistical cut-off points and may not necessarily represent targets for, or examples of, optimum nutritional status. Indeed, by definition, 2.3% of the children in the American reference population will fall below -2 z-score weight-for-height and be defined as acutely undernourished. Regardless of whether percent of median, centile, or z-score is used, if anthropometry is being used to determine who needs nutritional intervention, children falling below the selected cut-off point will be included in such an intervention and those falling above this cut-off point will not be included. In a population survey, the prevalence of undernutrition is calculated by dividing the number of children falling below the selected cut-off point by the total number of children measured. The cut-off point defining undernutrition is sometimes
determined using health outcome data. For example, adult undernutrition is
frequently assessed using the body mass index (BMI), which is also called the
Quetelet Index. It is calculated as the weight in kilograms divided by the
square of the height in meters (Wt/Ht |