Weight /Age Vs Weight/Height
Weight /Age Vs Weight/Height Michael Golden 21.04.99
Re: Weight /Age Vs Weight/Height Judith McGuire 21.04.99
Re: Weight /Age Vs Weight/Height Renuka Jayatissa 21.04.99
Re: Weight /Age Vs Weight/Height Tony Nelson 21.04.99
Re: Weight /Age Vs Weight/Height Rae Galloway 21.04.99
Weight /Age Vs Weight/Height Stephen Oppenheimer 22.04.99
RE: Weight /Age Vs Weight/Height Abe Parvanta 22.04.99
no title Abe Parvanta 22.04.99
Re[2]: Ngonut: Weight /Age Vs Weight/Height David Robinson 22.04.99
Weight /Age Vs Weight/Height Patricia Haggerty 23.04.99
Weight /Age Vs Weight/Height Andre M.N. Renzaho 26.04.99
Anthropometry Pierre Nabeth 03.05.99
Re:Anthropometry Mercedes de Onis 28.05.99

Date: Wed, 21 Apr 1999 12:36:48 +0100

From: Michael Golden <m.goldenatabdn.ac.uk>

Subject: Weight /Age Vs Weight/Height


Dear Mike ,

I would like to bring to your attention a problem I have faced in two recent trips to countries which would have been considered food secure (to a greater or lesser degree) until fairly recently and are currently experiencing emergency conditions in some areas, affecting sizeable proportions of the population.

Health care infrastructure is quite developed, though grossly under funded, and growth monitoring is a routine activity. Weight for age is the anthropometric indicator used. In both these countries stunting is prevalent, with 25-35% of the children affected.

My dilemma is twofold, one is, how to target children for therapeutic/ supplementary feeding programmes when Wt/Age is confounded by stunting; the other is, to asses the degree of nutritional risk of the population accepting that prevalence of acute malnutrition is often used as an outcome indicator, albeit a very late one of food insecurity.




Dear Anon,

Thank you for your question. There are many issues involved and I cannot give a short answer I'm afraid. However, this is a critical issue for all involved in undernutrition. As the management of stunting and wasting are completely different and weight/age mixes up the two conditions, weight-for-age cannot be used to decide on the correct intervention for a population or to determine needs, it cannot be used as a criterion for admission or discharge of an individual from a program, and the efficiency of that program cannot be judged. There is also a real danger of concentrating on the wrong population groups with an intervention. I know that most academics, UNHCR, WFP and WHO-nutrition division strongly advocate use of weight/height and height/age only, whereas many Ministries of Health, the IMCI initiative (WHO and UNICEF) and Growth Monitoring Programs still use weight/age (exclusively?) for nutritional assessment.


1. Part of the problem comes from the general use of the term "malnutrition" to mean different things to different people, and the nomenclature is a mess: it would be very helpful if we all used the same terms in the same way!

The past and present situation:

WFA was originally used in the Gomez classification of degrees of malnutrition WFA (and oedema) was also unfortunately used for the Wellcome classification - this has led to much confusion because this classification equates the term marasmus with a WFA of less than 60% of standard (and also oedematous malnutrition with kwashiorkor) - so that normally proportioned small children could then legitimately be termed marasmic which has all the connotations of wasting.

The road-to-health chart and growth monitoring component of GOBI also used WFA as the primary indicator of malnutrition.

Those looking after severely malnourished children have always advocated weight-for-height for monitoring their patients - it is not unusual for children who are less than 60% WFA to fulfill both the criteria for admission and discharge at the same time because they were severely stunted - so it is nonsense to use a WFA classification.

In 1973 one of Philip Jame's Ph.D. students (? B.Axel) did a survey of a small Caribbean Island and found a very high prevalence of "severe malnutrition" but when James examined the children they all appeared normal clinically. This was due to stunting and was presented in the student's Thesis and at a seminar in LSHTM. With this demonstration, John Waterlow popularized the critical difference in weight-for-age, weight-for-height and height-for-age in several papers and WHO publications in the mid and late 1970s. Since that time most of those doing surveys of nutritional status have differentiated stunting and wasting and over the next few years a W Keller from WHO produced statistics to show that stunting was the dominant reason for low weight/age, affecting almost 40% of children globally. Those running programs for "acute medical" care of "malnutrition" then used weight-for-height/oedema as the criteria for admission and ignored the stunted.

However, weight/age continued to be used for "public health monitoring" to give an overall picture of the state of nutrition and at an epidemiological level relates well to the mortality rate. Weight/Height being related to short term, and height/age to longer term mortality risk in populations and the mixed index weight/age, by combining the two risks, is a good discriminator for use in population health.

We know what to do about wasting, but do not really know what to do about stunting. Perhaps this is why in so many countries, with active growth monitoring and intervention, the prevalence of stunting has hardly changed (see below).

Other terms that have been used include: "Acute malnutrition" to denote wasting and oedema (but as low weight-for-height can be long standing this is not entirely appropriate) "Chronic malnutrition" to denote stunting. with the acute or chronic malnutrition being called moderate if below -2SD and severe if below -3SD

"Global malnutrition" to denote moderate plus severe wasting plus oedema (i.e.. including all except for stunting)

Clearly, we need to get the nomenclature clear and agreed upon - there are a great many types of malnutrition - as there are for cancer for example.

Oncologists would never lump either benign and malignant tumors, or skin cancer, leukaemia and breast cancer together to give an overall number because the prognosis and treatments are so different - why do we persist in doing an analogous lumping together for the forms of malnutrition?

I suggest that we "recapture" the term marasmus from the Wellcome classification and use it to exclusively denote severe wasting, and retain kwashiorkor for oedematous malnutrition so that marasmus/kwashiorkor could be the name applied to those who need treatment according to the new WHO manual.

I also suggest that we strictly differentiate wasting and stunting to indicate the degree and type of malnutrition that a person has.

and to drop the term "underweight" as a measure of weight/age as it is confusing; further we should not use the term "malnutrition", without qualification, in technical documents or statements.


2. The importance of making the distinction of Weight/Age and Weight/Height

The importance in making the distinction is that the time frame and management of wasting and stunting are completely different.

Those who are wasted are treated with well established protocols and regain their weight rapidly, usually within one month; there seems to be no long term sequelae of experiencing an acute episode of wasting. The treatment works and should be applied appropriately only to those who will benefit from it.

Those who are stunted, if of normal birth weight, have taken about half their life to fall two standard deviations and fulfill the criterion for stunting! The prenatally malnourished IUGR reach this stage much more quickly, but are much more resistant to treatment and probably should be classified separately.

The stunted will need to grow at an accelerated rate for at least half their age to make up the deficit - so a 6 month old will need at least 3 months of treatment and a two year old will need to be treated for a whole year - even in ideal circumstances this is the case (e.g. it took about 2 years for catch up in height of Swiss children with stunting due to coeliac disease after correct treatment started). Stunted children do not catch up spontaneously unless the household problems with the (micro)environment, care and diet given by the caretakers change permanently. Indeed, UNICEF's conceptual malnutrition framework applies mainly to stunting rather than wasting or kwashiorkor.

Short term feeding that is successful for wasting and oedema fails to address the problems of stunting and will not work. Stunting is associated with long term problems of impaired psycho social development.

If a child is of low weight-for-age because he is both stunted and wasted - we can correct the wasting with therapeutic feeding, but at what weight-for-age is his wasting corrected? We cannot tell from this information. A major teaching hospital I visited in East Africa use weight/age as the admission criterion and could not decide when to discharge a child because there was no target weight for the correction of wasting - they discharged some prematurely and others after they had "recovered".

As stunting is much more common than wasting there is problem of overwhelming numbers of children being referred for, and getting, short term therapeutic of supplementary feeding with most having no prospect of recovery in weight-for-age and no clear idea by the staff of when the child should be discharged and the whole program will fall into disrepute because the "prescribed treatment" is not working.

If a weight-for-age survey is done and from the demography the numbers of children with "malnutrition" is calculated, a program designed and supplies ordered as if all the "malnourished" are wasted (or oedematous children) who require short term therapeutic /supplementary feeding then there will be a very expensive excess provision by three or four fold. The supplies will expire before they can possibly be used for the correct purpose - so they will be either dumped or used incorrectly. Such experience may prevent programs being implemented to treat the truly wasted and oedematous on subsequent occasions and would be a travesty brought about simply by planners failing to appreciate the difference between weight/age and weight/height.

The rate of "falling behind", to become stunted, is sufficiently slow for a major famine to occur without much change in population height-for-age, and weight-for-age to be a very blunt instrument for detection of the need for emergency intervention.

Data (presented at ACCSCN) from Indonesia, now undergoing a crisis, shows that weight-for-age is much lower in the rural areas and is not particularly bad in the urban areas and is not changing substantially - whereas when weight-for-height is taken it is clear that there have been minor changes in the rural areas but a recent and large deterioration in the urban areas. If weight/age was used as the criterion then one may conclude that the present crisis has had little effect, there is little cause for concern and that resources should continue to go to traditional rural development programs. When weight/height is looked at we see that there is an urgent need for a program to be initiated in the urban areas if there is not to be a catastrophe.

For the IMCI program, where many children are referred to clinical facilities because they are "malnourished", it is likely that many children will get inappropriate treatment - indeed one training video produced by IMCI clearly shows a stunted child who is labeled as severely malnourished - the clinical facilities do not have the appropriate means to do anything to address stunting even if they recognised that this was the problem!


3. The perceived difficulty in taking height measurements.

Some think that taking height is difficult, time consuming and the length boards are often not available. This is not the experience of profession aid agencies working in very difficult emergency conditions. I would ask the doubters whether taking height is more difficult than organizing a cold chain for a vaccine or taking blood and measuring haemoglobin? And yet we do not question these other procedures because we need them to deliver the services or make the diagnosis. Why should we be satisfied with inadequate data for the diagnosis of malnutrition? Why are some advisors quite content to make the wrong diagnosis and give the wrong treatment? Is it because they believe that malnutrition always the same no matter how it is defined and all we have to do is to "feed the children and watch them grow"? Is it all because they think that this simple measurement routinely made by staff in refugee camps and under tremendous pressure, is to difficult to implement? And if it is really to difficult then we should switch to using MUAC, which is another measure of wasting, and has repeatedly been shown to be a better indicator of mortality risk than weight-for-height or weight-for-age.

Certainly measure weight/age - but that is the first step only. All children with a low weight-for-age should immediately have their height taken, wasting and stunting quantified and the appropriate treatment given.

Mike Golden

Date: Wed, 21 Apr 1999 14:21:49 +0100

From: Jmcguireatworldbank.org

Subject: Re: Weight /Age Vs Weight/Height



Three points are worth noting on this anthropometry discussion:

1. "Growth monitoring" (in quotes) should not be measuring status. Very few countries are actually doing growth monitoring or growth promotion. What they're doing is periodically measuring anthropometry. The should be measuring growth, i.e. weight gain. If programs looked at weight gain and actually designed their programs around weight GAIN we'd begin to prevent malnutrition (and prevent rests largely in empowering households to use their own resources to care for their children rather than kneejerk food distribution/supplementation). Using anthropometry to screen for feeding programs (whether for acute malnutrition or for supplementary feeding) is not growth monitoring/promotion.

2. The problem with height/llength (and he problem is most with length) is not that it's too difficult, but rather that the error is so high relative to the Std. Dev. that measured heights have little value. Errors are high for physicians and nurses, not just the auxiliary nurses, the real workhorses of the health systems. Because of the error, the value of height for growth monitoring/promotion is limited. It is also limited for screening. I use height, from periodic surveys not monitoring, as an indicator of overall food securitiy in a country.

3. Weight status is a pretty good indicator of nturitional status if you concentrate on kids under two years old (where we should be focusing most of the prevention work after we work on fetal malnutrition). It is less error prone relative to Std. Dev. than height. Except in emergencies/famines we shouldn't use wt/ht because we're essentially measuring acute disease rather than growth. We simply can't interpret wt/ht on national level and it varies a great deal by season.



Judith McGuire

LCSHD; I 7-207

World Bank, 1818 H St. NW, Washington, DC 20433

Phone: 202-473-3452, Fax: 202-522-1201

Email: jmcguireatworldbank.org

Date: Wed, 21 Apr 1999 14:18:43 +0100

From: Renuka Jayatissa <renuka.jayatissaatlshtm.ac.uk>

Subject: Re: Weight /Age Vs Weight/Height


Dear Prof. Golden,

It was an interesting discussion. In my country (Sri Lanka) there is a supplementation programme for all the children under 5's.

Weight/age is taken as a indicator to enter the supplementation programme. When the weight is corrected for the age or it is parallel to the standard growth curve the child is discharged from the programme after 6 months period. But some children never come above the redline of the growth chart. They will be automatically discharged from the programme after completing the 5th year. Is there anything to be done for those children? But we never check the height. It is a good point to remember about the height but the feasibility of checking of the height is low. It will be needed additional training and resources. We are so concerned about the cost.

Thanking you.


Dr. Renuka Jayatissa.

Date: Wed, 21 Apr 1999 14:41:19 +0100

From: Tony Nelson <tony-nelsonatcuhk.edu.hk>

Subject: Re: Weight /Age Vs Weight/Height


What Mike says makes a lot of sense and I am sure most of us would agree.

However I still have a concern that even if height is routine measured, some children will still be defined as "stunted" because they come from ethnic background A rather than ethnic background B. One only has to travel on the underground in Hong Kong to be convinced that the southern Chinese are physically much smaller than western Europeans! I wonder whether ethnic differences are taken sufficiently into account, when prevalence of stunting is assessed in South Asia and elsewhere?


Tony Nelson

From: rgallowayatworldbank.org

Date: Wed, 21 Apr 1999 10:11:55 -0400

Subject: Re: Weight /Age Vs Weight/Height


In response to Tony Nelson's comment, WHO recommends the National Center for Health Statistics because there is evidence that well-nourished children (under five) from high income groups in all countries can be compared to these standards. There are of course individual variations and that's why looking at growth, as Judy McGuire suggested in her email, is particularly important if we are to monitor the nutritional status of one child. Kids can be consistently growing along the 5th percentile and be perfectly healthy and normal. If a child has always been about 20th percentile but fallen to the 5th percentile, then that child needs attention! The stunting we see in developing countries is not, for the most part, from genetic differences. It is from differences in feeding small children. Secular changes we've seen with Japanese migrating the US show that after even one generation on a US diet, heights start to be similar to the rest of the population that made up the NCHS reference population. I would discourage countries from developing their own standards for this reason and because it eats up valuable resources and energy that are needed to counsel mothers (and continuously follow up with them) to adequately feed their children and treat disease.



Rae Galloway, Nutritionist

East Asia Human Development, The World Bank

1818 H Street, N.W. Room MC7-845, Washington, D.C. 20433

Tel. 202-473-7232, Fax. 202-522-3394


From: Stephen Oppenheimer <stephen.oppenheimeratpaediatrics.oxford.ac.uk>

Subject: Weight /Age Vs Weight/Height

Date: Thu, 22 Apr 1999 09:16:25 +0100


Dear Tony,

I agree with you, but there may be people who dismiss your 'underground observation' saying that the southern Chinese so-called 'ethnic stunting' you refer to in Hong Kong is merely a residual effect of former nutritional compromise and that this should be corrected in the population as a whole by a secular trend over the next generation. This does not appear to be the case. As Sophie Leung and Susan Lui showed in their detailed dietary/anthropometric cohort and crossectional studies, a) there is no direct evidence of systematic nutritritional compromise in Hong Kong children even when stratified by family income; b) there originally were secular trends in length growth patterns but these flattened off in Hong Kong some time ago; c) there is evidence of a stable systematic difference between northern and southern Chinese populations, the former being on average longer (hence larger) at all ages. The conclusions in both Sophie's and Susan's theses were that the marked divergences between growth patterns for Hong Kong children and NCHS standards particularly in length after 15 months, were likely to be a result of ethnic differences. The NCHS samples themselves were a mixed bag and hence flawed but this does not affect the conclusions on Hong Kong kids.

Stephen Oppenheimer


Stephen Oppenheimer DM, FRCP.

Green College, Oxford.

From: "Parvanta, Ibrahim" <ixp1atcdc.gov>

Subject: RE: Weight /Age Vs Weight/Height

Date: Thu, 22 Apr 1999 10:24:58 -0400


I would like to echo Rae Galloway's comment, and refer you to a recent publication (Mei, Z. et al. Improving trend of growth of Assian refugee children in the USA: Evidence to support the importance of environmental factors on growth. Asia Pacific J. Clin. Nutr (1998) 7:111-116). which presents nutrition surveillance data in the United States from 1979 to 1993, and compares the growth of S.E. Asian immigrant children to that of other ethnic groups in the country. The authors state, "The Asian refugee children under 5 years of age showed a progressive and significant decline in the prevalence of low height-for-age and low weight-for-age....... By 1993, the growth status of Asian refugee children was comparable with that of other ethnic groups. This marked improvement over a short period strongly suggests that the poor growth status earlier observed among recently immigrated Asian children was due to nutritional and health factors related to socioeconomic conditions, rather than to genetic factors, as is often suggested".


Abe Parvanta

Centers for Disease Control and Prevention Atlanta, U.S.A.

e-mail: ixp1atcdc.gov

From: "Parvanta, Ibrahim" <ixp1atcdc.gov>

Subject: Ngonut: RE:

Date: Thu, 22 Apr 1999 13:23:18 -0400


Dr. Jayatissa's concern about cost of anthropometric equipment and training is understandable. However, there is also a cost (and most likely much higher) for inadequate nutritional screening, rehabilitation and follow-up of children. Most likely, such cost comparisons have been done, and perhaps some colleagues on this listserve are aware of relevant studies and could share them.


Abe Parvanta

CDC, Atlanta, USA

e-mail: ixp1atcdc.gov

From: David Robinson robinsondatwho.ch

Date: Thu, 22 Apr 1999 16:55:14 +0100

Subject: Re[2]: Weight /Age Vs Weight/Height


Weight for height, as Mike says, is routinely used in refugee camps. With all their experience do the camp health staff really find it that difficult to produce a reliable figure that can actually guide the management of the child?

Or, as I suspect, are the problems mainly experienced by people who don't measure children regularly? Could someone working day-to-day in the refugee field tell us?


David Robinson

Date: Fri, 23 Apr 1999 17:51:18 -0400

From: haggertyatmacroint.com (Patricia Haggerty)

Subject: Weight /Age Vs Weight/Height



To begin to address the issues raised by Anon correctly, we need to distinguish between the two questions: first, how to target children for supplementary feeding; second, how to assess nutritional risk in a population.

The first question is about assessing individual nutritional status. Most supplementary feeding programs are designed to correct acute starvation, or respond to emergency situations. In these cases weight for height below a cutoff (eg. -2 SD) is a practical way to target the neediest children, because it responds quickly to feeding. "Monitoring" the progress of these kids (and the effectiveness of the program) can be done by tracking absolute weight gain or, even better, the rate of weight gain over time, or by measuring W/H at some point later in time and looking at the change.

Some supplementary feeding programs are part of integrated child health programs and aren't responding to acute situations. In these programs, choosing beneficiaries based on W/H cutoffs is really arbitrary and doesn't make much sense. Targeting is more difficult... program managers need to assess an array of factors, such as the number of kids under 5 in the home, birth spacing, age of the mother, education/literacy of the mother, access to health care, access to food, related poverty indicators, and weight for height, among others.

Measuring weight and height is preferable in most situations, and a lot of field people are doing it right, but its true that height is often difficult because it is logistically cumbersome and requires more training and resources than weighing. Moreover, a lot of height equipment used in the field is badly constructed, inaccurate, or used incorrectly. Also in many cultures measuring height isn't possible because of social beliefs and connotations about what's happening - In parts of rural central AFrica, for example, there's a belief that measuring a child's height means someone is building a coffin for her.

Your suggestion of MUAC is probably the best approach for targeting children for supplementary feeding programs where, for one reason or other, height is not practical. However, MUAC is less specific than W/H, so it will capture a wider net, i.e. it will identify more kids malnourished than W/H. Some folks argue that capturing a wider net wastes resources, but in most cases the differences are marginal and in my opinion that is splitting hairs over who is poorest among the poor. MUAC is very practical, very easy, and very cheap. MUAC can also be used to graduate recovered kids from the program because it responds quickly in the youngest kids. As kids approach 4 and 5 years MUAC is less reliable.

The other question, of assessing nutritional risk of a population, is a totally different issue. Here you have to be clear whether you're talking about "risk", or about nutritional status. The term risk is used loosely in our field. Risk is the likelihood that something may happen; a predictor. W/H is a not a measure of an individual's risk of acute malnutrition, it is acute malnutrition. But is is a measure of risk of death. The famine early warning people have developed some very sophisticated measures of risk of famine (which leads to acute malnutrition) that are not based on anthropometric indicators. H/A doesn't measure risk of stunting, but may predict low work productivity, poor school performance, etc.

If you're assessing the nutritional status of a population, the right choice of indicators depends on what you need the information for. Is it to get an overall picture of current health status to inform and advocate for nutrition? Or is it to identify problems, design interventions, influence nutrition policy, allocate resources? If its the former, weight for age is still a good, overall measure of the general health status of a population. It isn't a good indicator of an individual's status, nor of the nature of malnutrition in a population.

It is useful for advocacy, for getting the attention of high level officials and in many instances for making general cross country comparisons (also very powerful with government officials!). W/A isn't useful for designing program strategies or assessing change in a population over time.

Height for age is a good long term indicator of population nutritional status, and probably the best we have for assessing change in a population over time. H/A is also powerful policitically. Donors and decision makers use H/A more often these days to assess population nutritional status and change, but unfortunately have unrealistic expectations for how fast H/A will change. Many expect to see major changes in 5 years or less, but H/A change realistically is much slower than that. We need to be assessing 8, 9, 10 year windows and longer.

... but, of course, H/A alone is not sufficient if your assessment has the purpose of allocating resources to nutrition intervention strategies. For that, a variety of other indicators may be needed - H/A, W/H, iron status, vitamin A status, salt iodization, feeding practices, etc.

W/H is a good indicator at the individual level but shouldn't be used to assess overall nutritional status of a population. All the right reasons have been mentioned - it is affected by too many short term factors (epidemics, food shortage, natural disasters, season) and is very situation and location specific. W/H in a population survey has its place, however, in highlighting where trouble spots exist or, in demonstrating the congruence between acute malnutrition and mitigating factors, like diarrhoea, or food shortage, or illustrating the age at which children are most vulnerable. But even in countries where repeat surveys have occured, W/H change over time is not a valid comparison, unless the study is highly controlled for confounding, because the factors influencing the first measure are likely to be different than those influencing the latter measure.

The DHS program has been reporting the 3 measures of nutritional status (stunting, wasting, underweight) for almost 15 years, and distinguish the differences in interpretation of these indicators. Comparisons across countries are made using H/A and W/A only, not W/H. Many countries have now had repeat surveys, and typically only the changes in H/A are analyzed and reported, because the other indicators have the limitations discussed.

In conclusion, we should first distinguish what we're trying to assess and why, before we choose the indicators we're going to use. That would 1) narrow the choice of indicators tremendously; and 2) ensure that we really measure what we need to.


Patricia Haggerty

Child Survival Technical Support, Macro International Inc.

11785 Beltsville Drive, Calverton, MD 20705

Date: Mon, 26 Apr 1999 11:18:22 +0100

From: Andre Masumbuko Nzunzu Renzaho <amnratstudent.unimelb.edu.au>

Subject: Weight /Age Vs Weight/Height


Dear David Robinson,

Of course the weight for height is routinely used in refugee camps to monitor the progress of malnourished children under nutritional rehabilitation. The data generated by this indice are reliable depending on:

1. the type of measuring board being used (i.e.) adapted to the climate condition of the country where the intervention is being delivered (reference to physics)

2. the level of training of the anthropometric assistants (people who measure the child). If they did not receive adequate training, then the data they will produce are more likely to contain measurement errors

3. The same theory (2) is applied on the way children are weighed. It is very easy to underestimate or overestimate the weight of a child if training has not been offered to the anthropometric assistants.e.g the scale is hanged up to the height > the height of the person reading the weight etc.

Another concern would be the reference table for weight . This table reflect the situation of children in developed countries, but is being used in underdeveloped countries. Perharps NGOnuters colleagues could be able to provide you with sufficient info with respect to this matter.

For a good management of the child, the latter is measured twice a week and the time of weighing should be the same through out the treatment. I hope you would understand why. For example if the first weighing was at 8.00 A.M before any meal and the second is at 2.00 PM after a child had some meals, then there is a measurement error introduced ( weight fluctuation).Consequently the inference drawn from these two different weights could lead to false decisions.

Hope this helps.


Andre M.N. RENZAHO, Master of Public Health

The university of Melbourne, Australia

Date: Mon, 03 May 1999 12:37:26 +0100

From: pnabethattoptechnology.mr (pierre nabeth)

Subject: Anthropometry


I think that Mike's reply to the question concerning W/A vs W/H is complete and correct. I will just add that Weight for Age, Height for Age and Weight for Height are not ratios and that it is probably more appropriate to use W-A, H-A or W-H abbreviations.

But I will have some questions/comments on the contributions of other Ngonut(ers ?):


W-A vs W-H, MUAC

I agree that measuring W-A is a more appropriate index than H-A to monitor individual nutritional status in "usual situations". It is easy to collect; any deterioration of the nutritional status will affect weight and therefore, the W-A will decrease quite rapidly.

But I am not sure that measuring height is much more difficult than measuring weight or that the usual errors have a greater effect on indicators based on height than on weight. The use of an indicator based on weight is certainly more relevant than height alone as nutritional status deterioration occurs so much later with height than with weight that any intervention based on height measurement will be delayed.

However, in "unusual situations" such as emergencies, famines, refugee context, there is no other choice than measuring W-H: the individual health status can change in a few days and the weight has to be standardised with an accurate reference measure (age is rarely accurate) to plan timely programmes.

I do not understand why in non-acute situations, selection of beneficiaries for feeding programmes based on W-H cut-offs is arbitrary. W-H is commonly admitted as the best index of acute malnutrition and a child with a W-H index below –2 SD is considered as suffering of acute malnutrition (or wasting). I agree that in these situations, wasting should not be the only reason for admission to feeding programmes and that social considerations are also important, but this does not mean that basing admissions on W-H indicator is arbitrary.

MUAC can be more specific and more sensitive than W-H. Acute malnutrition is generally accepted as W-H below –2 SD but there is no consensus for MUAC. MUAC with a cut-off at 110 mm will probably detect less malnourished children (lower sensitivity) than W-H below –2 SD.

MUAC specificity and sensitivity do not depend only on the cut-off, but also on the population age-distribution: medium MUAC regularly increases from 150 mm to 200 mm between 6 months and 59 months of age (1). As a consequence, for a same cut-off , MUAC sensitivity will be higher (and specificity lower) in a younger population than in an older one; MUAC should probably not be used alone but in combination with another measure such age or height.

As P.Haggerty wrote, low W-H is not a risk of acute malnutrition but is considered as acute malnutrition. However, in crisis situations, children evolve quickly from normal nutritional status to moderate and then severe malnutrition. As a consequence, moderately malnourished children are considered to be at risk of severe malnutrition and later, of death.

Repetition of prevalence of malnutrition studies based on W-H seems to me as pertinent as repetition of studies based on W-A or H-A indexes.

The choice of the collected measures has nothing to do with the methodology but with the information you need. In many situations, measuring the evolution of prevalence of acute malnutrition (W-H < -2 SD) is essential to monitor feeding programmes and the easiest way is to conduct anthropometric nutritional surveys.


Ethnic differences

There are quite a lot of studies which indicated that at least until 5 years old, differences between individuals are more likely due to environmental conditions than to race (2), (3), … According to Merimee (4), even growth of pygmies is comparable to growth of other African children until 5 years old. Difference occurs later.


(1) National Center for Health Statistics. NCHS growth curves for children, birth to 18 years, United States. Rockville Md., NCHS, Vital and Health Statistics, 1977; 11 (165).

(2) Martorell R, Mendoza F, Castillo R. Poverty and stature in children in: Waterlow JC. Linear growth retardation in less developped countries. New-York: Raven Press, Nestlé, Nutrition workshop series 1988; 14: 57-73.

(3) Habicht JP, Martorell R, Yarbrough C, Malina RM, Klein RE. Height and weight standards for preschool children. How relevant are ethnic differences in growth potential ? Lancet 1974; iv: 611-15.

(4) Merimee IJ, Zapf J, Hewlett B et al. Insulin-like growth factors in Pygmies. New Engl J Med 1987; 316; 15: 906-911.


Pierre Nabeth, Medical Epidemiologist

Ministry of Health, Nouakchott


From: deonismatwho.ch Mercedes de Onis

Date: Fri, 28 May 1999 09:04:08 +0100

Subject: Re:Anthropometry


Dear Dr Nabeth,

With regards to your comment below about the possibility of using MUAC in combination with age or height, I like to mention that WHO and CDC have recently developed reference data for this purpose. The reference data can be found in the papers/book below. They should be easily accesible in the Ministry library, but if they are not please let me know and I will be happy to send you reprints.

Kind regards,


Mercedes de Onis, Department of Nutrition

World Health Organization


1.- de Onis M, Yip R, Mei Z. The development of MUAC-for-age reference data recommended by a WHO Expert Committee. Bulletin of the World Health Organization 1997;75:11-18.

2.- Mei Z, Grummer-Strawn LM, de Onis M, Yip R. Development of a mid-upper-arm circumference-for-height reference, including a comparison to other nutritional status screening indicators. Bulletin of the World Health Organization 1997;75:333-341.

3.- Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. Technical Report Series No. 854. Geneva: World Health Organization, 1995.