¹International Emergency and Refugee Health Branch, National Center for Environmental Health, Division of Emergency and Environmental Health Services, US Centers for Disease Control and Prevention, Atlanta, Georgia, USA

²RNIS Coordinator, UN ACC/Sub-committee on Nutrition, Geneva, Switzerland

Correspondence to: B A Woodruff, CDC Mailstop F-48, 4770 Buford Highway, Atlanta, GA 30030, USA. E-mail: Bwoodruff@cdc.gov

^aGuarantor: BA Woodruff.

^bContributors: BAW reviewed relevant biomedical literature, interviewed experts in the field, and wrote the first draft of the article. AD constructed overall design of the article, contributed major ideas, and revised subsequent drafts of the article.

^cCurrently with Save the Children, UK.

Objective: To outline the difficulties and suggest potential solutions in anthropometric assessment of adolescents during humanitarian emergencies.

Design: Literature review.

Setting: Multiple settings in which the nutritional status of adolescents has been assessed using anthropometric measurements.

Subjects: Adolescents in multiple populations.

Interventions: None.

Results: The use of anthropometry may be more difficult in adolescents than in other age groups because anthropometric indices in normally nourished adolescents change with age and sexual development. Moreover, survey and reference populations may differ in the age at which certain pubertal landmarks are attained, requiring adjustment for differences between survey and reference populations. Adolescent populations may also differ by ethnicity in various body proportions that affect anthropometric indices. Adjustment may be required when the body proportions of adolescents in the reference population differ from those in the population assessed.

Conclusions: Although no definitive recommendation can be made regarding which anthropometric indices are the most appropriate for adolescents, some revisions may improve current practices. Weight-for-height could be used for prepubertal adolescents and body mass index could be used for postpubertal adolescents. Because cut-off points are age-specific, age should be collected as accurately as possible for all adolescents measured during screening or survey activities. The WHO-recommended reference population of US adolescents is inappropriate in most populations of adolescents. Adolescents should never undergo nutritional assessment in isolation; other population subgroups should be included, and other health, nutrition and food data should be collected at the same time.

Sponsorship: The United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition.

European Journal of Clinical Nutrition (2002) 56, 1108-1118. doi:10.1038/sj.ejcn.1601456

adolescence; anthropometry; nutrition assessment; nutrition surveys; feeding programs

Introduction

In displaced and emergency-affected populations, the most common method of assessing the overall nutritional status in a population is to weigh and measure children 6-59 months of age (MSF, 1995; The Sphere Project, 1998; WHO, 1995, 2000). However, in the past few years, emergencies in Europe, Central Asia and Africa have highlighted the nutritional vulnerability of other population subgroups, such as elderly adults (Branca et al, 1996; CDC, 1992; Collins, 1993, 1996; McNabb et al, 1994; Toole & Serdula, 1996). Adolescents, defined by the World Health Organization (WHO) as persons 10-19 y of age, have not traditionally been considered at disproportionately elevated nutritional risk in emergency situations. Nonetheless, because of rapid growth in stature, muscle mass and fat mass during the peak of the adolescent growth spurt, some dietary requirements are as high or higher in adolescents than in other age groups (WHO, 2000). For example, the 2420 kcal required per day by adolescent 15-19 y of age is the highest energy requirement of any age group. The recommended general ration of 2100 kcal per person per day for populations wholly dependent on relief food is based on a distribution of age and sex which assumes that 20% of the population are adolescents 10-19 y of age and 56% are adults (WHO, 2000). In populations with a higher proportion of adolescents or adults, this ration may provide insufficient energy. In addition to energy needs, the requirement for many micronutrients, including vitamin A, thiamine, riboflavin, niacin, folic acid, vitamin B₁₂, vitamin C and iodine, reaches levels required by nonpregnant adults. Moreover, rapid growth produces a higher requirement among adolescents 10-14 y of age for calcium than any other population age group except pregnant women.

Although little is known about the short- and long-term effects of acute undernutrition during adolescence, this age group may also present a nutritional opportunity. In many cultures, a large proportion of girls have their first pregnancy during adolescence. Improvement in nutritional status before and during pregnancy can improve pregnancy outcomes, including maternal death, fetal death and preterm delivery, experienced by pregnant adolescents (ACC/SCN, 2000).

This paper discusses the assessment of undernutrition in adolescents in emergency situations. Because of the focus on humanitarian emergencies, this discussion will include only an evaluation of acute undernutrition; it will not address many other important nutritional problems in adolescents, such as anemia and other micronutrient deficiencies, poor nutritional habits, eating disorders and obesity.

This report will also not directly address chronic undernutrition. In adolescence, chronic undernutrition also delays normal maturation (ACC/SCN, 1992; Ahmed et al, 1998; Brabin et al, 1997; Delgado & Hurtado, 1990; Dreizen et al, 1967; Kurz & Johnson-Welch, 1994; Singh, 1995). Although chronic undernutrition is an important and widespread problem (ACC/SCN, 2000) with multiple adverse health outcomes, it is not usually the highest nutritional priority in emergency situations. In such situations, acute undernutrition is often a more pressing problem and, at least in young children, may account for a substantial proportion of overall mortality (Young & Jaspars, 1995). Moreover, the distinction between acute and chronic undernutrition among adolescents and adults is not as clear as it is among young children. In children and young adolescents, chronic undernutrition leads to stunting. Because adults and older adolescents no longer increase their height, they cannot become further stunted.

The WHO recommends using clinical criteria to screen adolescents for severe undernutrition to determine the need for admission to therapeutic feeding centres. To assess the prevalence of acute undernutrition in a population, the WHO currently recommends using body mass index (BMI)-for-age. Each measured adolescent's BMI is compared with BMI of members of the same sex and age in the US reference population. Adolescents falling below the fifth centile, or a locally defined cut-off point, are defined as malnourished. In addition, WHO recommendations describe a method to adjust, at least in part, for potential differences in the ages of maturation between the survey and reference populations (WHO, 1995).

Examples of past anthropometric surveys of adolescents

Since their publication, the WHO recommendations have been used to analyze the results of several surveys assessing the nutritional status of adolescents in both stable and displaced populations in developing countries. The survey population characteristics and the estimated prevalence rates of thinness and stunting are summarized in Table 1. When put into context, the results of these surveys cast doubt on the validity of the WHO-recommended method in accurately assessing the prevalence of protein-energy undernutrition in adolescents.

Two of the surveys included adolescents from families with relatively high socioeconomic status and still found substantial acute undernutrition among adolescents. Although no data were specifically collected on health outcomes, four of the surveys chose samples of adolescents from school. As a result, the survey samples included only adolescents who were well enough to attend school. Nonetheless, three of these surveys estimated that the prevalence of thinness (indicating undernutrition) was greater than 50%. In the three surveys in refugee populations that collected additional health data, no evidence existed of elevated morbidity or mortality in the general population (Cookson et al, 1998; Woodruff et al, 1998, 1999).

Monitoring in the refugee camps demonstrated that the amount of food contained in the general distribution generally was adequate. Moreover, no indications existed of substantial undernutrition in any other segment of the population. In all these populations, recent surveys of children less than 5 y of age had estimated a low prevalence of acute undernutrition. One would have to hypothesize extraordinarily inequitable intra-household food distribution to explain the high prevalence of acute undernutrition among adolescents if the survey results are valid.

In general, the prevalence of thinness does not exceed the prevalence of stunting. In five of the eight adolescent surveys that measured both, the estimated prevalence of thinness exceeded that of stunting by a factor of 1.6-6.3. Although little is known about the usual ratio of the prevalence rates of thinness and stunting among nutritionally compromised adolescents, these results would be highly unusual in children less than 5 y of age. An analysis of the results of 175 nutrition surveys from throughout the world demonstrated that, among children 12-23 months of age, stunting was 2.5-12.5 times more common that thinness, depending on the region where the survey was done (Victora, 1992). In addition, in the WHO Global Database on Child Growth and Malnutrition, only Fiji has a higher prevalence of thinness than of stunting (de Onis & Bloessner, 1997). Although the applicability of these data to adolescents is not precisely known, in adolescents, at least those who are prepubertal and still growing, the prevalence of stunting would be expected to exceed the prevalence of thinness because stunting is cumulative, while thinness results from relatively recent undernutrition and is reversible.

Complications of adolescent anthropometry

The anthropometric assessment of undernutrition among adolescents presents several problems that are not present in the assessment of young children.

Changes in body proportions with age

In young children, the index weight-for-height is assumed to be constant regardless of a child's age. That is, a well-nourished 3-y-old child is assumed to have the same 'ideal' weight as a stunted 5-y-old child of the same height. Although this assumption may not be entirely true, the lack of age data does not grossly interfere with the use of weight-for-height in children less than 5 y of age. In contrast, weight-for-height and BMI change with age in older children, adolescents and adults. In populations in industrialized countries, BMI reaches a nadir at about 6 y of age, then rises steadily until middle age (Cole, 1991; Hammer et al, 1991; Rolland-Cachera et al, 1991; Figure 1).

The Cormic Index (sitting height divided by standing height) measures the ratio of trunk length to stature and it declines throughout childhood because leg length increases faster than trunk length during prepubertal growth (Gerver & De Bruin, 1995). However, because the adolescent growth spurt is made up disproportionately of growth in the trunk, it produces a rise in Cormic Index in later adolescence (Gerver & De Bruin, 1995; Pathmanathan & Prakash, 1994; Yun et al, 1995). At least one study shows a sharp change in the Cormic Index with onset of the adolescent growth spurt (Yun et al, 1995). Other anthropometric measurements also change with age in adolescence. Mid-upper arm circumference (MUAC) rises progressively throughout adolescence at a rate greater than that in early childhood (Frisancho, 1981).

Because of these changes in anthropometric indices and measurements with age, persons from whom anthropometric measurements are taken must be compared with persons of the same age in reference populations. Unfortunately, persons in many developing countries do not know their own ages. In young children age can be approximated by asking a mother about significant events on a local calendar that coincided with the child's birth. Such a technique may be much more difficult when asking adolescents or parents about local events that occurred 10-19 y previously. In any case, using such techniques to ascertain age takes substantial time, both in preparation of a local calendar and in posing extra questions during data collection.

Pubertal development

Superimposed on the more gradual age-related changes are the more rapid changes in anthropometric indices and measurements which occur during sexual development. For example, during the adolescent growth spurt, the highest rate of weight gain follows the highest rate of height gain (Riley, 1990). This leads to an acceleration in the rate of rise of BMI shortly after reaching the peak velocity in the increase in height, and this rise in BMI is more related to pubertal development than to chronologic age (Riley, 1990; Riley et al, 1989).

Sexual development occurs at different ages in different populations (Eveleth & Tanner, 1990). As a result, differences between study populations and reference populations in sexual maturation can confound comparisons. For example, the comparison of prepubertal children in a study population with postpubertal persons of the same age in a reference population would be invalid because the normal BMI for these two groups would be quite different. Chronic undernutrition can also delay sexual maturity and the adolescent growth spurt (Ahmed et al, 1998; Delgado & Hurtado, 1990; Eveleth & Tanner, 1990) This nutrition-related delay can exaggerate differences in age of sexual maturation and the growth spurt between undernourished survey populations and well-nourished reference populations (Tanner, 1976).

Skinfold thickness and MUAC are also related to sexual development. The onset of puberty changes the rate of subcutaneous fat deposition and fat distribution, as measured by skinfold thickness (Deurenberg et al, 1990; Gasser et al, 1993; Martinez et al, 1993; Samsudin, 1990) and these changes occur independently of chronologic age (Garcia Llop et al, 1990). Similarly, MUAC changes with onset of puberty (Gasser et al, 1993; Samsudin, 1990).

Adjustment for differences between the survey and reference populations, as recommended by WHO, requires the collection of data on the age at which certain landmarks of sexual maturation occur in the survey population. Each subject must be assessed for pubertal changes; however, in practice, judging Tanner breast or genital stages during field surveys is difficult. To do so, health workers require extensive training and, in most field situations, exposing adolescents' breasts and genitals is not culturally acceptable. The single maturational landmark that might easily be collected in the field is age of menarche. In populations where adolescents know their ages, age at menarche can generally be recalled by girls and women and has been measured successfully in many surveys (Brabin et al, 1997; Chaturvedi et al, 1994; Delgado & Hurtado, 1990; Leung et al, 1996; Riley et al, 1989; Samsudin, 1990; Simondon et al, 1997). However, the situation is different for boys, for whom no such easily assessed landmark exists. Self-assessment of genital and breast stage, using photographs or line drawings, has been validated in some countries, but has not been fully evaluated in developing countries or in a variety of cultures (Duke et al, 1980; Hardoff & Tamir, 1993; Neinstein, 1982).

Ethnic differences in body size and shape

Although well-nourished, healthy children younger than 5 y of age are of similar size and shape worldwide (Habicht et al, 1974), school-age children and adolescents may not be. Older adolescents who have completed their growth spurt and have essentially become adults may reflect adult differences in body shape. For example, adults display differences in Cormic Index (Norgan, 1994a, b). Figure 2 demonstrates the substantial effect which differences in Cormic Index have on BMI. As a result, it may be inappropriate to compare older adolescents, who may manifest ethnic differences in anthropometric indices and measurements, with a single universal reference population.

Advantages and disadvantages of various anthropometric indices

Anthropometric indices for use in humanitarian emergencies must be calculated from measurements that require only inexpensive equipment, be simple to teach to relatively inexperienced survey workers, and be easy to interpret in the field. In addition, the ideal index used to assess adolescent nutritional status in emergencies would allow compensation for differences between the survey and reference populations in age, sexual development and ethnicity. Indices most commonly considered for use in emergencies include MUAC and combinations of weight and height. Skinfold thickness has been studied in adolescents and found to be a useful measure of body protein and fat stores, but it has not been widely considered for use in emergencies because it is more difficult to obtain accurate measurements in field conditions.

Weight-for-height measures

As described above, WHO recommends the use of BMI to assess adolescent nutritional status. To screen adolescents for admission to nutrition programs, other organizations, including Médecins sans Frontières (MSF) and Action Contre la Faim (ACF), use weight-for-height reference tables extended to older children and adolescents (Michael Golden, personal communication, January 2000). This method uses existing pediatric reference tables of weight-for-age and height-for-age to calculate the median weight and 70% of median weight for adolescents of each sex and height category.

All weight-for-height indices have several drawbacks when employed to assess the nutritional status of adolescents. No data exist that directly correlate weight-for-height, BMI or the Rohrer Index (weight in kg)/(height in m³) with functional or health outcomes in adolescents. As described above, all indices using weight and height are correlated to some extent with age (Cole, 1991; Figure 1). Although the association between age and BMI is not as strong as that between age and weight-for-height, the use of BMI still requires the collection of accurate ages and ages of maturational landmarks from each subject (Figure 1). The Rohrer Index may be less age-dependent during adolescence than other indices that combine weight and height. However, some studies have found that the Rohrer Index is correlated with height, especially among older adolescents (Foster et al, 1977; Freedman & Perry, 2002; Omura et al, 1993).

The height and weight measurements required to assess weight-for-height, BMI and Rohrer Index may be difficult to obtain during an emergency. During severe famines where adolescents are affected, many of the most severely undernourished who require admission to therapeutic feeding centers cannot stand, making measurement of height impossible. Many studies have reported that gross weakness and flexor contractions prevented measurements of weight or height in a substantial proportion of severely undernourished adults (Collins et al, 2000). Moreover, the necessary equipment, including scales and height boards, may not be available. The calculation of BMI and Rohrer Index may be unfamiliar to field workers trained to evaluate weight-for-height in young children.

Regardless of the specific index used, evaluation of nutritional status must take into account the presence or absence of edema. Many adolescents and adults develop edema when severely undernourished (Collins, 1993; Collins & Myatt, 2000), which artificially increases an individual's weight, resulting in a weight-for-height, BMI or Rohrer Index appearing more normal than would be expected given the degree of emaciation. In addition, because adolescents and adults with edema have a poorer prognosis than those who are equally undernourished but do not have edema, identification of adolescents with edema should be a priority.

Because of the extra weight of the fetus, other products of conception, and added maternal tissue and fluid, indices using weight and height may not accurately indicate the nutritional status of pregnant adolescents. During pregnancy, other measures, such as weight gain during pregnancy or MUAC, must be used to judge nutritional status.

MUAC

MUAC is relatively simple and easy to measure and has been recommended for use in rapid screening of adults for undernutrition to determine the need for admission to feeding programs (Collins, 1996; Collins et al, 2000; Ferro-Luzzi & James, 1996; James et al, 1994). In many well-nourished adult populations, a reasonable correlation exists between MUAC and BMI. However, no data exist that directly correlate MUAC with other measures of body fat and under-nutrition, such as BMI, in adolescents. In addition, no data relate MUAC cut-off points with functional or health outcomes in adolescents. As a results, no validated cut-off points defining undernutrition in adolescents have yet been established. Although the published results of several nutrition assessment surveys of adolescents in less-developed countries include MUAC data, the lack of a reference population and cut-off points has prevented the formulation of estimates of the prevalence of undernutrition from these data (Chaturvedi et al, 1994; Delgado & Hurtado, 1990; Hussain et al, 1996; Owa & Adejuyigbe, 1997; Qamra et al, 1990; Simondon et al, 1997; Singh, 1995; Spurr et al, 1983; Strickland & Ulijaszek, 1992; Toriola, 1990)

Similar to indices using weight-for-height, MUAC changes substantially with age during adolescence, as shown in adolescent populations in developed countries (DHHS, 1981; Frisancho, 1981). As a result, a different cut-off point must be used for adolescents of different ages, necessitating collection of an accurate age for each subject to determine his or her appropriate cut-off point defining undernutrition.

MUAC also changes with sexual development. The rapid addition of soft tissue which occurs with puberty results in a more rapid rise in MUAC at this time than before or after puberty. One study clearly demonstrates a greater MUAC in postmenarcheal female adolescents than premenarcheal female adolescents of the same age (Simondon et al, 1997).

In spite of the convenience and ease of measurement, MUAC measurement requires careful training and supervision to prevent wrapping the measuring tape too tightly or too loosely and thus introducing measurement error. One study estimated that the smallest change over time detectable in MUAC was 8-10% when measurements were taken by different observers (Bishop & Pitchey, 1987). A second study demonstrated that MUAC measurements show more interobserver variability than weight and height measurements (Fuller et al, 1991).

Potential solutions

No single method has yet proven adequate for assessing undernutrition in adolescents in emergency situations. Below are listed some possible strategies for overcoming the major difficulties. However, none of these strategies has been fully validated, and all need further investigation.

Changes in body proportions with age

An anthropometric index that is relatively less related to age is required if adolescents do not know their ages. The Rohrer Index may offer such advantages over weight-for-height, BMI or MUAC.

Pubertal development

Simpler methods of determining pubertal stage may, in the future, allow adjustment for this complicating factor. Such methods could include appropriate line drawings or photographs of different Tanner stages. In contrast, it may be easier and more accurate when assessing adolescents to consider prepubertal and postpubertal adolescents separately. Ideally, this would require the assessment of the presence or absence of only one landmark of sexual development, rather than the determination of different levels of a development indicator, such as Tanner breast or genital stages. Such separation may make it possible to include prepubertal adolescents with school-aged children and postpubertal adolescents with adults when choosing which anthropometric indicator and cut-off points to use.

Interethnic differences in genetic growth potential

A method has been proposed to account for at least part of the differences in body shape in adults by calculating a BMI that is adjusted for the Cormic Index (Norgan, 1994a, b, c; Norgan & Jones, 1995). However, this procedure remains untested in adolescents (Nicholas Norgan, personal communication, January 2000). Theoretically, such an adjustment could also be applied to other anthropometric indices calculated from weight and height, such as weight-for-height or Rohrer Index. The normal changes in Cormic Index throughout adolescence may complicate the application of this technique to adolescents. Moreover, some data indicate that chronic undernutrition changes the Cormic Index (Malina, 1991). Stunted children may have a greater Cormic Index than comparable children without stunting.

An additional method of compensating for ethnic differences is to choose different, more appropriate, cut-off points to define undernutrition when using a reference population whose ethnicity differs from that of the survey population. Some populations may be genetically thinner than the reference population, requiring a lower-cut-off to define undernutrition. The WHO recommends such a procedure when using anthropometry to screen persons for admission to nutritional rehabilitation (WHO, 1995). Nonetheless, the procedure of determining appropriate cut-off points would require either data on health outcomes to estimate functional cut-off points or data on a well-nourished population of similar ethnicity to derive statistically defined cut-off points. Collection of such data for multiple ethnic groups would be expensive and time-consuming.

Using a local reference would permit comparison of a survey population to well-nourished adolescents of the same ethnicity. Although this has been used in published surveys (Chamruengsri et al, 1991), creating a reference population is a difficult task, requiring substantial resources. Separate reference populations cannot be created for each nationality or ethnicity. An alternate possibility would be to create a few reference populations for use with major ethnic categories. Investigators could then choose the reference that most closely matches the survey population. For example, if a survey population has an average Cormic Index of 0.50, the investigators would use a reference population with a similar average Cormic Index to match the study population with a reference population with a similar body habitus. Of course, such a strategy would make comparison of results from different surveys difficult if these surveys used differ-ent reference populations to calculate the prevalence of undernutrition. In addition, it would require measuring sitting height as well as standing height in all surveys of adolescents.

A single international reference population consisting of adolescents from multiple countries could be used, similar to that currently under development for children less than 5 y of age. A reference population for use in measuring overweight and obesity was described in recently published work. This reference includes children and adolescents from national surveys in six countries (Cole et al, 2000). Cut-off points for BMI defining overweight and obesity were identified by determining the centile among 18-y-old adolescents in each national survey population that matched the adult BMI cut-off points of 25 and 30. That centile was then applied separately for males and females to each one-half-year age group from 2 to 17.5 y to determine the cut-off BMI value at for each age and sex group. The cut-off points for each national population were then averaged to produce age- and sex-specific cut-off points for the combined international reference. The same procedure could be used to determine cut-off points that correspond to the adult cut-off points for various degrees of undernutrition, ie the BMI values of 16, 17 and 18.5. This proposed reference and the method of determining cut-off points for BMI eliminate the bias that may result from using as a reference a sample from a relatively obese population, such as adolescents from the United States. However, its use still requires accurate ages because the BMI cut-off points change substantially with age. In addition, the reference proposed may not be appropriate for populations with different body shape, such as Nilotic Africans and Australian aborigines.

Use of other data

When assessing adolescents, additional data from the same population could be collected. Such data could include the prevalence of undernutrition in young children, the prevalence of undernutrition in adults, rates of morbidity and mortality, and information on food security, food distribution and alternative sources of food. In most situations adolescents are unlikely to be the only population group with substantial undernutrition; therefore, a discrepancy between adolescents and other population groups may indicate problems with the assessment of adolescents.

Unless anthropometric measures are valid, that is, they truly measure nutrition and health status, they may not be useful at all. Other measures, such as strength or other functional outcomes, may better reflect an individual's risk for nutrition-related morbidity or mortality. Experience in screening adults for admission to therapeutic feeding programs has demonstrated that clinical signs (for example, apparent dehydration and edema) predict mortality better than BMI (Collins & Myatt, 2000; Collins et al, 2000). Such an approach, although as yet untested in adolescents, may be able to distinguish adolescents in need of therapeutic feeding from those who do not need such intensive intervention. However, clinical signs may not be useful for measuring the overall prevalence of lesser degrees of undernutrition in a population because persons with less severe undernutrition may not be so markedly impaired nor exhibit such a distinctive clinical picture.

Measures of muscle function, such as grip strength, shuttle run or maximum jump height, may be able to detect moderate degrees of wasting. One author proposes that muscle performance may actually be affected earlier in the development of undernutrition than body composition (Jeejeebhoy, 1994). However, not all studies show a relation between nutritional status and muscle function (Benefice et al, 1999; Shizgal et al, 1986). Much work is needed on this area before such measures can be used in the field.

Future research needs

Certain questions need to be answered before the best method of assessing nutritional status in adolescents can be definitively recommended.

Anthropometric index

The association of various indices; such as weight-for-height, BMI and Rohrer Index, with age and height should be further explored in adolescents in less-developed countries. Such exploration should include analysis of existing data from past surveys that measured weight, height, MUAC and age of adolescents. These re-analyses should include surveys undertaken in populations with a variety of health conditions and various degrees of undernutrition. In addition to investigating the association between anthropometric indices and age and height, analyses of data from future surveys could include estimates of the error induced in the estimated prevalence of undernutrition by various degrees of uncertainty about age.

Defining functional cut-off points

Longitudinal studies are required to determine whether adolescents falling below specific cut-off points for weight-for-height, BMI and Rohrer Index have elevated morbidity or mortality, poor pregnancy outcome, suppressed growth or decreased work ability or physical performance measures. These studies should be conducted in a variety of situations among adolescents with different levels of undernutrition.

Markers of pubertal development

The accuracy of self-reported Tanner stage for breast and genital development among adolescents needs to be tested in a variety of populations. For example, adolescents' self-assessment upon viewing drawings or photographs of different Tanner stages could be compared with physical examination findings. Studies could also explore the use of other markers of sexual development, such as the extent of axillary hair, which may be more culturally acceptable. Moreover, the ages at which various landmarks of sexual development are achieved should be described in many different populations, both normally nourished and undernourished, to determine which markers can be used for adjusting for different developmental ages between survey and reference populations.

Adjusting for differences in body shape

Surveys should explore the utility of adjusting anthropometric indices, such as BMI, Rohrer Index or others, for differences in body shape in distinct populations.

Practicality of measurements and calculations

The practicality of obtaining various measures should be explored in field situations. Survey organizers should assess the ease of training survey workers in measuring MUAC, weight and height, as well as assessing inter-observer variability in these measures when assessing adolescents.

Vulnerability

More nutrition surveys should include assessments of the nutritional status of different age groups to determine the relative vulnerability of young children, adolescents, women of child-bearing age, elderly and other population subgroups in different types of humanitarian emergencies.

Use of anthropometry at all

Other measures, such as strength, clinical signs or other functional outcomes, should be explored as indicators of adolescent nutritional status by conducting longitudinal studies and comparing the results of these measures to the results of anthropometric measurements. Much more work needs to be done to determine whether appropriate measures of muscle function exist, to describe a reference population and cut-off points, and to determine the sensitivity and specificity of proposed methods to detect moderate and severe undernutrition.

Conclusions and recommendations

A fundamental dilemma exists measuring adolescent undernutrition: a system simple enough for nonexpert field staff is required for use in emergencies, but the reality of adolescent undernutrition is complex. Until better tools for nutritional assessment of adolescents are developed, current recommendations regarding the use of anthropometry in adolescents must be critically examined. Their use may yield misleading results that stimulate inappropriate interventions. No standard method of anthropometric assessment of the nutritional status of adolescents can be recommended at this time. However, recent developments have suggested potential techniques that can be used until the results of the research described above are available and a specific method can be recommended. The following components should be included in any anthropometric assessment of adolescents.

Screening for severe undernutrition

Until better methods can be developed and validated, screening for severe undernutrition to determine the need for therapeutic feeding should use clinical and social criteria as recommended for adults (Collins et al, 2000).

Correction for differences in age of sexual maturation

Some measure of the age of attaining specific pubertal landmarks should be measured during nutrition surveys of adolescents. One measure that may be useful includes age of menarche in females. Unfortunately, validated markers that are practical for field use do not exist for males. Correction for differences between the survey population and the reference population should be undertaken if the necessary data are available for the reference population used.

Prepubertal adolescents

Because young adolescents may be more similar to children, the most appropriate index for use in measuring undernutrition prevalence among prepubertal adolescents may be weight-for-height, at least until other indices are more fully investigated. Individuals can be compared to an existing reference using the preliminary weight-for-height tables as described above. Data from this original reference population should be recalculated to provide direct weight-for-height cut-off points.

Postpubertal adolescents

Because older adolescents may be more similar to adults, BMI should be used until other indices have been more fully investigated.

Age

Both weight-for-height and BMI are age-dependent. Therefore, when these indices are used, age must be collected as accurately as possible for each person screened. It may be necessary to construct a local calendar to determine ages and to investigate the existence of systematically biased reporting of age that could lead to substantial underestimation or overestimation of the prevalence of undernutrition when weight-for-height or BMI are used.

Reference populations

The reference population of US adolescents, recommended by WHO for use with BMI, should not be used. The multinational reference population described by Cole et al (2000) may the best reference currently available for use in the less-developed world. Investigators should use the method for creating age-specific cut-off points described by Cole to determine cut-off points, in both percentage of median and Z-scores, which correspond to the adult BMI cut-off points of 16, 17 and 18.5, and these cut-off points should be widely disseminated. Use of other reference populations, especially those composed predominantly of ethnic groups not represented in the references mentioned above, should be tested.

Additional population groups

Given the lack of validated anthropometric procedures, adolescents should not undergo nutritional assessment in isolation. Such assessments should also incorporate evaluation of the nutritional status of other population subgroups, such as young children, women of child-bearing age, adults and elderly persons. A large discrepancy between the estimated level of undernutrition in adolescents and that in other population subgroups should stimulate investigation of the validity of the methods and results of the adolescent assessment.

Additional data

Nutritional assessment of adolescents should include, in addition to anthropometric measurements, clinical assessment of adolescents, the assessment of food availability and uses, the level of morbidity and mortality among adolescents and adults, and the presence and prevalence of micronutrient deficiencies.

Comparison of surveys

To assess the methods and comparability of surveys, all survey reports should describe in detail the anthropometric index used, how measurements were taken, which reference population was used, how individuals were compared to this reference, the cut-off points used to define various degrees of undernutrition, and any other ancillary data collected on the population of interest.

Acknowledgements

We are grateful to the following reviewers for their comments: Monika Blössner (WHO), Tim Cole (Institute of Child Health), Mary Corbett (Concern), Mercedes de Onis (WHO), Mike Golden (University of Aberdeen), Saskia van der Kam (Médecins sans Frontières), Carlos Navarro-Colorado (Action contre la Faim), Nick Norgan (University of Loughborough), Claudine Prudhon (Action contre la Faim), Simon Strickland (London School of Hygiene and Tropical Medicine), and Stanley Ulijaszek (Oxford University).

ACC/SCN. (1992). Second report on the world nutritition situation: volume I, global and regional results. Geneva: Administrative Committee on Coordination¾Subcommittee on Nutrition.

ACC/SCN. (2000). Fourth report on the world nutrition situation: nutrition throughout the life cycle. Geneva: United Nations Administrative Committee on Coordination¾Subcommittee on Nutrition.

Ahmed F, Zareen M, Khan MR, Banu CP, Haq MN, Jackson AA. (1998). Dietary patterns, nutrient intake and growth of adolescent school girls in urban Bangladesh. Pub. Health Nutr., 1: 83-92.

Benefice E, Fouere T, Malina RM. (1999). Early nutritional history and motor performance of Senegalese children, 4-6 y of age. Ann. Hum. Biol., 26: 443-455.

Bishop CW, Pitchey SJ. (1987). Estimation of the mid-upper arm circumference measurement error. J. Am. Diet. Assoc., 87: 469-473.

Brabin L, Ikimalo J, Dollimore N, Kemp J, Ikokwu-Wonodi C, Babatunde S, Obunge O, Briggs N. (1997). How do they grow? A study of south-eastern Nigerian adolescent girls. Acta Paediatr., 86: 1114-1120.

Branca F, Burkholder BT, Hamel M, Parvanta A, Robertson A. (1996). Health and nutrition survey of internally displaced and resident population of Azerbaijan¾April 1996. Atlanta, GA: Centers for Disease Control and Prevention.

CDC. (1992). Nutritional needs surveys among the elderly¾Russia and Armenia, 1992. Morbid. Mortal. Weekly Rep., 41: 809-810.

Chamruengsri K, Kietduriyakul V, Pava-Ro U, Sapsamarnwong C, Chandumpai P. (1991). Nutritional status of low socioeconomic school children at Srakaew Temple, Ang Thong. J. Med. Assoc. Thai., 74: 24-29.

Chaturvedi S, Kapil U, Bhanthi T, Gnanasekaran N, Pandey RM. (1994). Nutritional status of married adolescent girls in rural Rajasthan. Indian J. Pediatr., 61: 695-701.

Cole TJ. (1991). Weight-stature indices to measure underweight, overweight, and obesity. In: Anthropometric Assessment of Nutritional Status ed. JH Himes, pp 83-111, New York: Wiley-Liss.

Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. (2000). Establishing a standard definition for child overweight and obesity worldwide: international survey. Br. Med. J., 320: 1240-1243.

Collins S. (1993). The need for adult therapeutic care in emergency feeding programs. Lessons from Somalia. JAMA, 270: 637-638.

Collins S. (1996). Using middle upper arm circumference to assess severe adult malnutrition during famine. JAMA, 276: 391-395.

Collins S, Myatt M. (2000). Short-term prognosis in severe adult and adolescent malnutrition during famine: use of a simple prognostic model based on counting clinical signs. JAMA, 284: 621-626.

Collins S, Myatt M, Duffield AE. (2000). The assessment of adult nutritional status in emergency-affected populations. Geneva: ACC/SCN.

Cookson ST, Woodruff BA, Slutsker L. (1998). Prevalence of anemia and low body-mass index among adolescents 10-19 y of age in refugee camps in Dadaab District, Kenya. Atlanta, GA: Centers for Disease Control and Prevention.

de Onis M, Bloessner M. (1997). WHO global database on child growth and malnutrition p 710, Geneva: WHO.

de Onis M, Dasgupta P, Saha S, Sengupta D, Blössner M. (2001). The National Center for Health Statistics reference and the growth of Indian adolescent boys. Am. J. Clin. Nutr., 74: 248-253.

Delgado HL, Hurtado E. (1990). Physical growth and menarche in Guatemalan adolescents. Arch. Latinoam. Nutr., 40: 503-517.

Deurenberg P, Pieters JJ, Hautvast JG. (1990). The assessment of the body fat percentage by skinfold thickness measurements in childhood and young adolescence. Br. J. Nutr., 63: 293-303. MEDLINE

DHHS. (1981). Basic data on anthropometric measurements and angular measurements of the hip and knee joints for selected age groups 1-74 y of age, United States, 1971-1975. Hyattsville, MD: US Department of Health and Human Resources.

Dreizen S, Spirakis CN, Stone RE. (1967). A comparison of skeletal growth and maturation in undernourished and well-nourished girls before and after menarche. J. Pediatr., 70: 256-263. MEDLINE

Duke PM, Litt IF, Gross RT. (1980). Adolescents' self-assessment of sexual maturation. Pediatrics, 66: 918-920.

Eveleth PB, Tanner JM. (1990). Worldwide Variation in Human Growth 2nd edn Cambridge: Cambridge University Press.

Ferro-Luzzi A, James WP. (1996). Adult malnutrition: simple assessment techniques for use in emergencies. Br. J. Nutr., 75: 3-10.

Foster TA, Voors AW, Webber LS, Frerichs RR, Berenson GS. (1977). Anthropometric and maturation measurements of children, ages 5 to 14 y, in a biracial community¾the Bogalusa Heart Study. Am. J. Clin. Nutr., 30: 582-591. MEDLINE

Freedman DS, Perry G. (2000). Body composition and health status among children and adolescents. Prev. Med., 31: S34-S53.

Frisancho AR. (1981). New norms of upper limb fat and muscle areas for assessment of nutritional status. Am. J. Clin. Nutr., 34: 2540-2545. MEDLINE

Fuller NJ, Jebb SA, Goldberg GR, Pullicino E, Adams C, Cole TJ, Elia M. (1991). Inter-observer variability in the measurement of body composition. Eur. J. Clin. Nutr., 45: 43-49. MEDLINE

Garcia Llop LA, Ramada Benedito A, Rodriguez-Estecha P. (1990). Relationship between sexual maturation and subcutaneous fat. An. Esp. Pediatr., 33: 349-353.

Gasser T, Ziegler P, Kneip A, Prader A, Molinari L, Largo RH. (1993). The dynamics of growth of weight, circumferences and skinfolds in distance, velocity and acceleration. Ann. Hum. Biol., 20: 239-259.

Gerver WJ, De Bruin R. (1995). Relationship between height, sitting height and subischial leg length in Dutch children: presentation of normal values. Acta Paediatr., 84: 532-535.

Habicht JP, Martorell R, Yarbrough C, Malina RM, Klein RE. (1974). Height and weight standards for preschool children. How relevant are ethnic differences in growth potential? Lancet, 1: 611-614.

Hammer LD, Kraemer HC, Wilson DM, Ritter PL, Dornbusch SM. (1991). Standardized percentile curves of body-mass index for children and adolescents. Am. J. Dis. Child., 145: 259-263. MEDLINE

Hardoff D, Tamir A. (1993). Self-assessment of pubertal maturation in socially disadvantaged learning-disabled adolescents. J. Adolesc. Health, 14: 398-400.

Hussain A, Lindtjorn B, Kvale G. (1996). Protein energy malnutrition, vitamin A deficiency and night blindness in Bangladeshi children. Ann. Trop. Paediatr., 16: 319-325.

IRC. (1997). Nutritional status of school aged children in Kakuma refugee camp. Nairobi: International Rescue Committee.

James WP, Mascie-Taylor GC, Norgan NG, Bistrian BR, Shetty PS, Ferro-Luzzi A. (1994). The value of arm circumference measurements in assessing chronic energy deficiency in Third World adults. Eur. J. Clin. Nutr., 48: 883-894.

Jeejeebhoy KN. (1994). How should we monitor nutritional support: structure or function? New Horiz., 2: 131-138.

Kurz KM. (1996). Adolescent nutritional status in developing countries. Proc. Nutr. Soc., 55: 321-331. MEDLINE

Kurz KM, Johnson-Welch C. (1994). The nutrition and lives of adolescents in developing countries: findings from the nutrition of adolescent girls research program. Washington, DC: International Center for Research on Women.

Leung SS, Lau JT, Xu YY, Tse LY, Huen KF, Wong GW, Law WY, Yeung VT, Yeung WK, Leung NK. (1996). Secular changes in standing height, sitting height and sexual maturation of Chinese¾the Hong Kong Growth Study, 1993. Ann. Hum. Biol., 23: 297-306.

Malina RM. (1991). Ratios and derived indicators in the assessment of nutritional status. In: Anthropometric Assessment of Nutritional Status ed. JH Himes, pp 151-171, New York: Wiley-Liss.

Martinez E, Devesa M, Bacallao J, Amador M. (1993). Subscapular/ triceps index: percentile values in Cuban children and adolescents. Arch. Latinoam. Nutr., 43: 199-203.

McNabb SJ, Welch K, Laumark S, Peterson DE, Ratard RC, Toole MJ, Farley TA. (1994). Population-based nutritional risk survey of pensioners in Yerevan, Armenia. Am. J. Prev. Med., 10: 65-70.

MSF. (1995). Nutrition Guidelines 1st edn Paris: Medécins sans Frontières.

Neinstein LS. (1982). Adolescent self-assessment of sexual maturation: reassessment and evaluation in a mixed ethnic urban population. Clin. Pediatr. (Phila.), 21: 482-484.

Norgan NG. (1994a). Interpretation of low body mass indices: Australian aborigines. Am. J. Phys. Anthropol., 94: 229-237. MEDLINE

Norgan NG. (1994b). Population differences in body composition in relation to the body mass index. Eur. J. Clin. Nutr., 48: ((Suppl 3)) S10-25, discussion S26-27. MEDLINE

Norgan NG. (1994c). Relative sitting height and the interpretation of the body mass index. Ann. Hum. Biol., 21: 79-82.

Norgan NG, Jones PR. (1995). The effect of standardising the body mass index for relative sitting height. Int. J. Obes. Relat. Metab. Disord., 19: 206-208. MEDLINE

Omura M, Zinno S, Harada T, Inoue N. (1993). Evaluation of validity of five weight-height obesity indices. Fukuoka Igaku Zasshi, 84: 305-310. MEDLINE

Owa JA, Adejuyigbe O. (1997). Fat mass, fat mass percentage, body mass index, and mid-upper arm circumference in a healthy population of Nigerian children. J. Trop. Pediatr., 43: 13-19.

Pathmanathan G, Prakash S. (1994). Growth of sitting height, subischial leg length and weight in well-off northwestern Indian children. Ann. Hum. Biol., 21: 325-334.

Qamra SR, Mehta S, Deodhar SD. (1990). Evaluation of anthropometric indices for the assessment of nutritional status in girls¾III. Ind. Pediatr., 27: 1275-1279.

Riley AP. (1990). Dynamic and static measures of growth among pre- and postmenarcheal females in rural Bangladesh. Am. J. Hum. Biol., 2: 255-264.

Riley AP, Huffman SL, Chowdhury AK. (1989). Age at menarche and postmenarcheal growth in rural Bangladeshi females. Ann. Hum. Biol., 16: 347-359.

Rolland-Cachera MF, Cole TJ, Sempe M, Tichet J, Rossignol C, Charraud A. (1991). Body Mass Index variations: centiles from birth to 87 y. Eur. J. Clin. Nutr., 45: 13-21. MEDLINE

Samsudin. (1990). The age of menarche as related to physical growth of junior high school girls in Jakarta. Paediatr. Indones., 30: 254-260.

Shizgal HM, Vasilevsky CA, Gardiner PF, Wang WZ, Tuitt DA, Brabant GV. (1986). Nutritional assessment and skeletal muscle function. Am. J. Clin. Nutr., 44: 761-771.

Simondon KB, Simon I, Simondon F. (1997). Nutritional status and age at menarche of Senegalese adolescents. Ann. Hum. Biol., 24: 521-532.

Singh R. (1995). Secular increase in body size and nutritional anthropometric measurements of Indian children. In: Essays on Auxology ed. R Hauspie, G Lindgren & K Falkner, pp 332-333, Avon: Castlemead.

Spurr GB, Reina JC, Barac-Nieto M. (1983). Marginal malnutrition in school-aged Colombian boys: anthropometry and maturation. Am. J. Clin. Nutr., 37: 119-132.

Strickland SS, Ulijaszek SJ. (1992). Energy nutrition of Iban in Song and Kanowit, February-April 1990. Sarawak Museum J., 43: 135-196.

Tanner JM. (1976). Population differences in body size, shape and growth rate. A 1976 view. Arch. Dis. Child., 51: 1-2.

The Sphere Project. (1998). Humanitarian Charter and Minimum Standards in Disaster Response 1st edn Geneva: IFRC.

Toole MJ, Serdula M. (1996). Determining the nutritional status of the elderly in post-cold war Russia. Am. J. Public Health, 86: 299-301.

Toriola AL. (1990). Anthropometric assessment of nutritional status of Nigerian children. Trop. Geogr. Med., 42: 337-341.

Victora CG. (1992). The association between wasting and stunting: an international perspective. J. Nutr., 122: 1105-1110. MEDLINE

WHO. (1995). Physical status: the use and interpretation of anthropometry Report of a WHO Expert Committee Geneva: World Health Organization.

WHO. (2000). The Management of Nutrition in Major Emergencies 2nd edn, p 236, Geneva: World Health Organization.

Woodruff BA, Slutsker L, Cook ST. (1998). Prevalence of anemia and low body-mass index in adolescents 10-19 y of age in Kakuma camp, Kenya. Atlanta, GA: Centers for Disease Control and Prevention.

Woodruff BA, Duffield A, Blanck H, Larson MK, Pahari S, Bhatia R. (1999). Prevalence of low body mass index and specific micronutrient deficiencies in adolescents 10-19 y of age in Bhutanese refugee camps, Nepal, October 1999. Atlanta, GA: Centers for Disease Control and Prevention.

Young H, Jaspars S. (1995). Nutrition, disease and death in times of famine. Disasters, 19: 94-109.

Yun DJ, Yun DK, Chang YY, Lim SW, Lee MK, Kim SY. (1995). Correlations among height, leg length and arm span in growing Korean children. Ann. Hum. Biol., 22: 443-458.

Figure 1 Weight-for-height (W/S), BMI (W/S2), and Rohrer Index (W/S3) as a function of age during childhood and adolescence, constructed from medians of National Center for Health Statistics reference population (from Cole, 1991).

Figure 2 Body mass index as a function of Cormic Index in a hypothetical 70 kg, 175 cm man (from Norgan, 1994a,b,c).

Table 1 Summary of the results of surveys which used WHO recommendations to assess adolescent nutritional status