Posted December 1999
Special: Population, poverty and environment
Anthropometric, health and demographic indicators in assessing nutritional status and food consumption
by Simon Chevassus-Agnès
FAO Food and Nutrition Division
with the collaboration of Alain Marcoux, FAO Population Programme Service
Approximately 790 million people in the developing world subsist on diets that are deficient in energy . About 200 million children suffer from malnutrition and 2 billion people suffer from a variety of micronutrient deficiencies. The vast majority of the food-insecure, whether their malnutrition is due to deficiencies in energy or in micronutrients, live in low-income developing countries and mainly in the poorest areas of those. These numbers are only estimates, since no direct data are actually available to assess the magnitude of these problems, identify the causal factors thereof, or describe the nutritional status of the poor.
At the World Food Summit (WFS) in 1996, governments and heads of state committed themselves to reducing the number of undernourished people to half its current level (then estimated at 840 million) by the year 2015. In order to monitor progress towards this goal, the WFS called for better information to identify the food-insecure and vulnerable groups, assess the extent of low food intake and undernutrition, and ascertain the main causes contributing to food insecurity and vulnerability to malnutrition. Better information is expected to facilitate the identification of appropriate policy and programme responses and the targeting of interventions to improve food security and nutrition.
With this objective, governments undertook to define, develop and periodically update a Food Insecurity and Vulnerability Information and Mapping System (FIVIMS), regrouping data on food-insecure and vulnerable people - what socio-economic categories they belong to, where they are located, and what degree of food deficiency and undernutrition prevails among them. National data will be integrated into international databases collaborating in this exercise. The latter will be linked through a common, decentralized system for posting and disseminating information, allowing to better monitor food security and nutrition trends at the global level.
At the national level, the WFS called for close attention to monitoring the availability and nutritional adequacy of food supplies, particularly in areas at high risk of chronic or seasonal food insecurity and among nutritionally vulnerable groups. It also called for mechanisms to be established for collecting information on the nutritional status of the poor and on members of vulnerable and disadvantaged groups at sub-national levels. This note presents relevant methods in this context and their population dimensions.
2. Possible causes of low food consumption and poor nutritional status
Low or inadequate food consumption - and consequent poor nutritional status - may be the result of a variety of causes, often operating in combination. Figure 1 illustrates the array of relevant factors. It shows that key determinants of food consumption are food availability (whether from local production or other sources) and people’s access to that food (i.e. their capacity to produce or purchase). In addition, aspects related to the stability of supplies over time (obtaining a steady flow of food from an inherently irregular production) are essential. The stability of access to food (e.g. stable income) is important in terms of food security.
The above determinants are in turn affected by the socio-economic and political environment at large, in particular by economic conditions at the macro and micro levels (importance and characteristics of the agricultural sector; trade relations; livelihood systems; etc.). Educational levels and cultural values also play a role in shaping food habits, consumption patterns and food supply systems in general.
At the individual level, the degree of efficiency with which the body utilizes the food consumed is a key determinant of nutritional status. That efficiency is generally affected by poor health conditions and reduced significantly by specific diseases (infections, intestinal parasites, diarrhoea etc.).
3. Assessing nutritional status in a population
A variety of methods are commonly used for assessing the nutritional status of populations based on anthropometric, clinical and biochemical measurements. Although methods based on dietary characteristics do not directly measure nutritional status, they are very often used by nutritionists in the field to assess it. These methods are described below and the main indicators recommended in the context of FIVIMS are presented in Table 1.
Table 1. Core indicators for nutritional status
Children under 5 years: W/A, W/H, H/A (to assess the satisfactory character, or otherwise, of corporal growth)
Children 5-10 years: W/A, W/H, H/A (idem)
Adults (>20 years): BMI (to assess underweight and overweight)
Birth weight: percent under 2.5 kg at birth (to assess the nutritional situation of new-borns and of mothers during pregnancy)
Prevalence of night blindness (hemeralopy) in children (to assess the vitamin A nutritional status)
Prevalence of goitre [total goitre rate, TGR] (to assess the iodine nutritional status)
Prevalence of nutritional anaemia among women and pre-school children (to assess the iron and/or folates nutritional status)
Figure 1. Determinants of food consumption and nutritional status: a conceptual framework
Anthropometric measurements (body dimensions and composition) are often used as proxies for assessing the eventual extent and severity of malnutrition. The classical indicators in this respect have to do with the growth of children and body composition of adults. The most commonly used measurements are the body weight, height, age and sex of each individual, which allow to calculate the following indicators:
Weight-for-age (W/A) reflects body mass relative to chronological age
Low W/A is influenced by both the height of the child (height-for-age) and his or her weight (weight-for-height). Its composite nature makes interpretation complex. For example, weight-for-age fails to distinguish between short children of adequate body weight and tall, thin children. However, in the absence of significant wasting (see next paragraph) in a community, similar information is provided by weight-for-age and height-for-age as both reflect the long-term health and nutritional experience of the individual or population. In general terms, the world-wide variations and age distribution of low W/A are similar to those of low height-for-age.
Weight-for-height (W/H) reflects body mass relative to height
Low W/H (wasting or thinness) indicates in most cases a recent and severe process of weight loss, which is often associated with acute starvation and/or severe disease. However, wasting may also be the result of a chronic unfavourable condition. Provided there is no severe food shortage, the prevalence of wasting is usually below 5 percent, even in poor countries. The Indian subcontinent, where higher prevalence rates are found, is an important exception. A prevalence exceeding 5 percent is alarming given a parallel increase in mortality that soon becomes apparent. On the severity index, prevalence rates between 10-14 percent are regarded as serious, and above or equal 15 percent as critical. Typically, the prevalence of low weight-for-height shows a peak in the second year of life. Lack of evidence of wasting in a population does not imply the absence of current nutritional problems: stunting and other deficits may be present.
Overweight is the preferred term for describing high W/H. Even though there is a strong correlation between high W/H and obesity as measured by adiposity, greater lean body mass can also contribute to high weight-for-height. On a population-wide basis, high W/H can be considered as an adequate indicator of obesity, because the majority of individuals with high W/H are obese. Strictly speaking, the term obesity should be used only in the context of adiposity measurements, for example skinfold thickness.
Height-for-age (H/A) reflects height relative to chronological age
Low H/A is called stunting. Stunted growth reflects failure to reach linear growth potential as a result of sub-optimal health and/or nutritional conditions. On a population-wide basis, high levels of stunting are associated with poor socio-economic conditions and increased risk of frequent and early exposure to adverse conditions such as illness and/or inappropriate feeding practices. Similarly, a decrease in the national stunting rate is usually indicative of improvements in overall socio-economic conditions of a country. The world-wide variation of the prevalence of low H/A is considerable, ranging from 5 percent to 65 percent among the less developed countries. In many such settings, prevalence starts to rise at the age of about three months; the process of stunting slows down at around three years of age, after which mean heights run parallel to the reference curve. Therefore, the age of the child modifies the interpretation of the findings: for children below 2-3 years, low H/A probably reflects a continuing process of "failing to grow" or "stunting"; for older children, it reflects a state of "having failed to grow" or "being stunted". It is important to distinguish between the two related terms, length and stature: length refers to the measurement in recumbent position, the recommended way to measure children below 2 years of age or less than 85 cm tall; whereas stature refers to standing height measurement. For simplification, the term height is used to cover both measurements.
Body Mass Index
(BMI)= weight (in kg) ¸
[height (in m) squared] reflects also body mass relative to height and is mainly used for adults and adolescents. High BMI permits to assess degrees of overweight and obese people and low BMI to assess different levels of thinness (and of chronic energy deficiency).
At the level of an entire population, the information gathered through the above indices provides a basis for assessing the prevalence of unsatisfactory conditions under one or the other of the relevant criteria. In practice, the indicators will then be:
- For children under age five, the Z-score classification system is used for population-wide assessments including surveys and nutrition surveillance . For consistency with clinical screening, prevalence-based data are commonly reported using cut-off values, usually at minus two and plus two Z-scores from the median in the reference population . This implies that slightly more than 2 percent of the reference population will be classified as "malnourished" even if they are truly "healthy" individuals with no growth impairment .
- Prevalence of underweight children is the percentage of children with a weight that is more than two Z-scores below the referenced weight-for-age.
- Prevalence of stunted children is the percentage of children with a height that is more than two Z-scores below the referenced height-for-age.
- Prevalence of wasted children is the percentage of children with a weight that is more than two Z-scores below the referenced weight-for-height.
- For adults (active, usually 20-49 years), although reference data are less validated than for children, recommendations for threshold to be used are:
- for thinness:
grade 1: BMI 17.0 - 18.49 (mild thinness)
grade 2: BMI 16.0 - 16.99 (moderate thinness)
grade 3: BMI <16.0 (severe thinness)
- for overweight:
grade 1: BMI 25.0 - 29.9 (mild overweight)
grade 2: BMI 30.0 - 39.99 (moderate overweight)
grade 3: BMI ³
40 (severe overweight)
As a first step in the implementation of the FIVIMS, these indicators are presented in concise and analytical reports called Nutrition Country Profiles (NCP), prepared by FAO in collaboration with the countries. Mapping is used in the National Country Profiles to illustrate the main nutritional problems affecting the population. Selected maps from the NCP of India, illustrating the main indicators used to assess the nutritional situation in a country:
- Map 1 (PDF, 2.7KB)deals with the prevalence of underweight among children under five years of age. A prevalence rate above 30 percent is considered as very high; in effect, except Nagaland (29.5 percent, high), Mizoran (23.4 percent, high) and Meghalaya (13.0 percent, medium) all the states are in a severe situation.
- Map 2 (PDF, 2.7KB)deals with the prevalence of stunting among children under five years of age. Prevalence above 40 percent is considered as very high; here again, all states except Goa (low prevalence), Manipur and Meghalaya (medium) and Mizoram (high) have a very high prevalence of stunting.
- Map 3(PDF, 2.7KB) deals with the prevalence of wasting among children under five years of age. This indicator, very often used to assess acute malnutrition, is very sensitive and may change very quickly. When its value is greater than or equal to 15 percent, the situation of children is critical; between five and nine percent it is poor; and between 10-14 percent it is serious. Meghalaya and Himachal Pradesh show a much better situation than other states. However, for all states in orange the situation for this indicator is less critical than for the two preceding ones.
- Map 4 (PDF, 2.9KB) deals with the prevalence of a Body Mass Index inferior to 18.5 and with differences between men and women on this score. The results are in total agreement with children's indicators and enable some explanation for the origin of malnutrition between food intake and health conditions.
- Map 5 (PDF, 3.1KB) deals with the prevalence of vitamin A deficiency among children under six years of age and show that generally the situation is worse in the northern part of India with some exception .
The same indices listed above for children under five can also be calculated for children between ages five and ten. This is less commonly done although it also represents very useful information. Finally, measurements are also taken for adolescents (ages 11-17) and for adults above 50 years, but such measurements are difficult to use because the reference values are less well established for these age groups.
Anthropometry-based nutritional assessment has the advantage of being a universally applicable, inexpensive and non-invasive method. This procedure also is applicable to large sample sizes. It can be used to identify target groups of population or areas for intervention, as a tool for nutritional surveillance, and in cross-sectional evaluation.
However, the nutritional status of children alone is not in itself a good indicator of the nutritional status and of food inadequacy (an important explanatory factor of nutritional status) of a population, for two reasons. At the macro level, children under age 10 account for 22 percent of the total population in the developing regions as a whole and 11 percent in the developed ones. Their energy requirements, even in areas with a young age structure, represent at most 15 percent of the requirements of the total population. At the micro level, intra-household food allocation is not necessarily balanced (because of socio-cultural factors that influence the patterns of distribution), so the children’s nutritional status does not necessarily reflect that of the household as a whole.
Clinical signs of some disorders associated with micronutrient deficiencies are also used to describe and quantify the nature and extent of malnutrition; the most usual ones are listed in Table 1. Vitamin A deficiency is detected and measured through assessing the prevalence of night blindness; iodine deficiency, through assessing the prevalence of goitre; and iron deficiency, through assessing anaemia by measuring haemoglobin concentration.
The sources of the data used in NCPs are selected on the basis of their quality. Demographic surveys - in particular the Demographic and Health Surveys (DHS), which cover numerous developing countries - are one of the main sources of anthropometric data on mothers and children. These surveys have allowed preparing maps showing the prevalence of malnutrition ion the world at national and sub-national level. These surveys should be encouraged and the measurements of other age and sex groups (children above five, adolescents and the elderly) should be carried out in order to enrich the knowledge of the causal factors of malnutrition when necessary.
4. Measurement of food supply and consumption
The purpose of nutritional epidemiology is to evaluate the impact of food consumption (macronutrients and micronutrients) on the health status of the population. All the existing approaches to evaluating food consumption are based on the same principle, i.e. the comparison of energy and nutrient intakes with the estimated requirements of the population. Those approaches are presented below.
4.1 The food inadequacy approach
The food inadequacy approach developed by FAO attempts to assess the number of people and proportion of the population with inadequate food consumption by comparing a country’s Dietary Energy Supply (DES) obtained from the FAO Food Balance Sheets (FBS) to the energy requirements of individuals.
Since the DES is an average, estimating the proportion of the population that is underfed  requires some notion of the inequality in distribution of that supply within the population. This method also necessitates a good estimate of the minimum per person dietary energy requirements in the population considered. This in turn requires information on body stature and on levels of physical activity in the population considered.
Although this method might provide a good proxy of food consumption at national level, its validity is limited by the inaccuracies of food balance sheets and by the difficulties in obtaining a correct estimation of energy requirements of populations. Moreover, this method only provides national averages and does not provide any information on intra- and inter-household inequalities in food consumption.
4.2 Food consumption measures
The methods most commonly used to assess food intake are described briefly below. They can be classified in two major groups. Those of the first group provide a measure of food consumption and include the weighing method, the food record method, the 24-hour recall method and the food frequency method. Those of the second group provide an estimate of food consumption using income/expenditure/budget surveys.
4.2.1 Food consumption measures
This method is usually applied at the household level but can also be used for individuals. Food items to be used are weighed before the preparation of each meal, while information on the food consumed outside the home is obtain by interviewing each member of the household. This method is very accurate but cannot be used for long periods because it is time-consuming and expensive. Therefore, it cannot measure food consumption over time.
Food record method
In this method an individual regularly records the type and quantity of all food consumed. This method is accurate but requires very careful training and supervision by the survey taker and the close co-operation and motivation of the respondents for the task to be carried out consistently.
24-hour recall method
In this method, the respondent is asked to recall his/her food intake during the last 24-hour period. The interviewer records a detailed description of all foods and beverages consumed. This is most appropriate for assessing average intakes of foods and nutrients for large groups of individuals. This method is suitable at individual level but cannot be carried out for long periods of time because it is too demanding for the respondent.
Food frequency questionnaire
In this method, the interviewer records the frequency of the consumption of some specific food items associated with the phenomenon that needs to be described. This is the choice method in nutritional epidemiology, mainly in the field of non-communicable diseases.
4.2.2 Estimates of food consumption: income/expenditure/budget surveys
This method enables estimating food consumption from an assessment of the household food expenditures. As with the FAO food inadequacy approach, this method, which may be precise at household level, does not inform on the intra-household distribution of food.
Table 2. Indices of food supply and consumption
Food balance sheets (national level)
Dietary energy supply per person (total and major food groups) as a proxy of food intake
Proportion of energy provided by animal products.
Proportion of energy provided by proteins (and quantities supplied).
Proportion of energy provided by lipids (and quantities supplied).
Proportion of animal proteins in total protein.
Proportion of animal fats in total lipids.
Food consumption indices (national, sub-national, urban, rural)
Per person total energy intake
Contribution of the major food groups to energy intake, when available
Average intake of essential nutrients (protein, vitamin A, iron, iodine etc.)
Map 6 (PDF, 3.0KB) provides an illustration of the use of the first indicator above. It shows the average energy intake (kilocalories per consumption unit per day) and the respective contributions of protein, fat and carbohydrates in energy intake. A consumption unit (CU) corresponds to the energy requirements of a referenced man aged 20-39 years weighing 60 Kg, doing sedentary work. The energy coefficients of all the other individuals in the household are calculated proportionately based on energy requirements according to age, sex, physiological status and activity pattern. Based on the nutrient requirements and recommended dietary allowance for India (1990) the energy requirement of the CU is 2,425 Kcal.
5. Health and demographic indicators
Certain demographic and health indicators are very useful to better understand some aspects of malnutrition, either at the national level or for smaller geographic entities as required by the specific policy question at hand. The indicators most commonly used in this context relate to the following:
- Population size and density. These characteristics have an indirect influence on general aspects and characteristics of the food supply system - its size, spatial extension, complexity - and the magnitude of possible food problems to be tackled.
- Age and sex distribution. This is needed to assess the size of critical groups (children under five, the elderly, women of childbearing age etc.). Also, age and sex structures heavily affect the dependency ratio.
- Fertility. It determines the numbers of pregnant and lactating women, a nutritionally critical category. It also has an influence on household size, which in turn is correlated with some nutritional indicators.
- Urban-rural population distribution. Energy expenditure varies significantly between rural and urban residents. And, of course, their respective food supply systems differ markedly.
- Proportion of agricultural population in total population. This is a measure of a population’s dependence on its own agricultural resources (vs. foreign trade) for subsistence.
- Life expectancy - the main indicator of general health conditions at the population level.
- Infant and child (0-4 years) mortality rates. These are more specific indicators of health conditions for a sensitive sub-population. They often give leads regarding nutritional conditions, especially around weaning.
- Morbidity rates for important diseases - diarrhoea, measles, acute respiratory infection, AIDS.
6. Socioeconomic indicators
Socio-economic indicators are useful to gain some understanding of societal conditions that are known to affect people’s access to food. Those most commonly used to document the said conditions are::
- Gross domestic product (GDP) per person. ()This is the best single indicator of average living standards in a population, but - leaving aside inherent difficulties of estimation and error margins - its main limitation is that a given average level of per person income may give rise to quite different nutritional outcomes depending on the degree of inequality in income distribution.
- Proportion of people below the poverty line. This indicator addresses the above drawback and attempts to assess the size of the population which, because of its economic conditions, has limited and irregular access to food. (The poverty line, i.e. the minimum income required to cover basic needs, obviously depends much on the cost of securing basic food requirements.)
- Human Development Index (HDI). This composite index combines data on GDP, education and health so as to also encompass some non-economic dimensions of well-being and "development". The HDI adds valuable dimensions in international comparisons; when the matter is assessing country situations, however, it is best to resort to specific data from the social sectors deemed relevant. Information on educational levels, for instance, often is useful, especially as it relates to women, who usually are in charge of many nutrition- and health-related functions in the household.
In order to assess whether food consumption in populations is sufficient (quantitatively) and adequate (qualitatively), whether at the global level or at the country level, a range of methodological approaches is available. Ideally, one would want to know - for a sufficiently long period of time - the magnitude of food intake among and within households of a given country and the resulting nutritional status of the population. Knowledge of health status also is important, as that status is one of the major determinants of the nutritional situation at the population level. In this context, this document has presented a minimal set of indicators that can be used to develop these types of knowledge. The said indicators purport to provide both [i] the information basis for the required assessments and [ii] insights into the commonly identified causes of food insecurity and of the degradation of nutritional status.
1. FAO. 1999. The state of food insecurity in the world 1999. Rome.
2. The Z score measures the degree of dispersion of a series of observations (here W/A, H/A and W/H) in relation to the centre (here, the median) of the series.
3. "Reference" means that the individual observations are compared with an international standard describing satisfactory corporal development. The standard was developed from observations made in a multiethnic population in good health.
4. For details, see http://www.who.int/nutgrowthdb/intro_text.htm.
5. Map 6 will be commented upon subsequently.
6. FAO. 1996. 6th World Food Survey. Rome.