International Journal of Contemporary Pediatrics

Mehta N et al. Int J Contemp Pediatr. 2019 Mar;6(2):522-526

http://www.ijpediatrics.com pISSN 2349-3283 | eISSN 2349-3291

DOI: http://dx.doi.org/10.18203/2349-3291.ijcp20190525
Original Research Article

Comparison of midupper arm circumference and weight-for-height z

score for assessing acute malnutrition in children aged
6-60 months: an analytical study
Nirali Mehta1*, R. G. Bhatt2, Hetal Vora3, Bharat Parmar3

1
Pharma-Stats, Ahmedabad, Gujarat, India
2
Department of Statistics, School of Sciences, Gujarat University, Ahmedabad, Gujarat, India
3
Department of Pediatrics, Shardaben General Hospital, Smt. NHL Medical College, Ahmedabad, Gujarat, India

Received: 27 January 2019

Accepted: 05 February 2019

*Correspondence:
Mrs. Nirali Mehta,
E-mail: [email protected]

Copyright: © the author(s), publisher and licensee Medip Academy. This is an open-access article distributed under
the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial
use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT

Background: In clinical settings, wasting in childhood has primarily been assessed with the use of a weight-for-
height z score (WHZ), and in community settings, it has been assessed via the mid upper arm circumference (MUAC)
with a cutoff <115mm for severe wasting and 115-125mm for moderate wasting. Our recent experience indicates that
many wasted children were not identified when these cutoffs for MUAC were used.
Methods: Authors determined the cutoffs for MUAC to detect wasting in Indian children aged 6-60 mo. A secondary
analysis was carried out on data from 1446 children aged 6-59 mo. The area under the receiver operating curve was
used to indicate the most appropriate choice for cutoffs that related MUAC with WHZ. The MUAC measurement of
each subject was taken using standard technique. Following the World Health Organization (WHO) age and sex-
specific cut-off points, nutritional status of children was determined.
Results: The mean±SD age for the entire group was 19.8±13.6 mo, MUAC was 132±13mm, and 45% of subjects
were girls. Age-stratified analyses revealed that, for ages 6-24 mo, MUAC cutoffs were <120mm for a WHZ <-3 and
<125mm for a WHZ <-2 with a sensitivity of 68.3% and 64.7%, respectively, and a specificity of 82.6% and 83.4%,
respectively; for ages 25-60 mo, MUAC cutoffs were <135mm for a WHZ <-3 and <140mm for a WHZ <-2 with a
sensitivity of 63.7% and 65.4%, respectively, and a specificity of 81.6% and 78.3%, respectively.
Conclusions: The respective cutoffs for MUAC to better capture the vulnerability and risk of severe (WHZ <-3) and
moderate (WHZ <-2) wasting would be <120 and <125mm for ages 6-24 mo, <135 and <140mm for ages 37-60 mo.

Keywords: Midupper arm circumference, Wasting, Weight-for-height z score

INTRODUCTION of mortality up to six times in diarrhoea and acute

respiratory tract infections.2,3 To accomplish the UN
Malnutrition contributes to almost two-thirds of Millennium Development Goal 4, which aims to further
worldwide mortality, directly or indirectly caused by reduce under-5 mortality, it is essential to curtail child
diarrhea, pneumonia, measles and other infections among deaths occurring as a significance of malnutrition. 4 The
children under 5 years of age.1 In hospitalized Indian first step in this path will be to categorize and manage the
children, malnutrition has been shown to increase the risk set of malnourished children at risk for imminent death.

International Journal of Contemporary Pediatrics | March-April 2019 | Vol 6 | Issue 2 Page 522
 Mehta N et al. Int J Contemp Pediatr. 2019 Mar;6(2):522-526

The WHO has definite the severe acute malnutrition conducted using standardized monitoring and assessment
(SAM) in 6-60-month-old children as a weight-for-height of relief and transitions methodology which aims to
<-3 SD (severe wasting) of the reference population. 5 The estimate the prevalence of wasting among children aged
assortment of weight-for-height Z-score (WHZ) over 6-60 months.
other anthropometric criteria, namely weight-for-age Z-
score (WAZ) or height-for-age Z-score (HAZ), is based All surveys used two‑stage cluster sampling where the
on the fact that it has been exposed to be an indicator not probability of being sampled was proportional to the
only for nutritional status but also involves measurement population size. For each survey, the sample size was
of height that can be exploit to evaluate past nutritional calculated using emergency nutrition assessment (ENA)
status. Though, the use of WHZ for identification of software which was sufficient to estimate the wasting
SAM is associated with some inherent consequence, prevalence with a precision of ±5%. Informed consent
especially in the emergency setting. Difficulties arise was taken from all the households that were included in
owing to the inability to accurately weigh or measure the study.
length in sick children; the non-availability of standardize
weighing scales and height boards; the need for reference Equipment of global standard was used for
charts at all times; and the complex calculations to derive anthropometric assessment. For measuring height,
and interpret WHZ. Further, WHZ is a statistically weight, and MUAC, wooden infanto‑cum‑stadiometer,
derived parameter which depends on the nutritional status SECA 874 digital weighing scales and standardized
of the population from which the Z-score has been MUAC tapes were used, respectively.
derived. Recognizing these operational difficulties, in
2009 the UN endorsed mid-upper arm circumference Data on weight, height, MUAC, gender, and age for a
(MUAC) <11.5cm to be an age- and sex-independent total of 1466 children <5 years were used in this study.
diagnostic criterion for SAM, alongside WHZ<-3.5 The data cleaning was done by deleting the records of
However, MUAC-based and WHZ based malnutrition children <12 months, >60 months, and with missing data.
diagnoses have shown deprived correlation and the At individual study level, outliers were removed as the
children acknowledged as SAM based on one criterion ENA flags children who had WHZ score <-3SD or
are often missed if diagnosed using the other.6 Therefore, >+3SD from the survey mean. Survey mean is mean
the difficulty arises of the choice of MUAC and WHZ as WHZ in each survey. Post dataset cleaning, 1466 children
the criterion for selecting hospital admission among were included for the final analysis.
severely malnourished children in resource-poor
countries. The WHZ <-3 SD only means those cases who were
SAM by WHZ only and their MUAC were ≥115mm. The
Several studies have established the superiority of MUAC <115 mm only means those cases who were
MUAC over other anthropometric indices like weight- SAM by MUAC only and their WHZ were ≥-3 SD. The
for-age, height for- age, weight-for-height, WAZ and overlapping SAM cases were those children who were
HAZ as a criterion to predict mortality among under-5 SAM based on both WHZ <-3 SD and MUAC <115mm.
African children in the community as well as among
hospitalized African children.7,8 However, very few The raw data were rechecked, transferred to an SPSS file
studies have directly compared MUAC<11.5cm with (SPSS Inc.), and analyzed with the use of SPSS software
WHZ<-3 for predicting child mortality.8 In India, among (version 20; SPSS Inc.). To ensure consistency, z scores
all GAM cases, 54.9% children were diagnosed with for all anthropometric data were calculated in relation to
WHZ <-2 only, 11.7% with MUAC <125mm only, and the WHO growth standard via WHO Anthro software
about 33.4% children were identified with both criteria. (version 2.0.2, 2007; WHO). Children were classified as
This overlap of prevalence based on WHZ and MUAC moderately wasted when the WHZ was <-2 SD and as
varies between countries and also within countries.9 severely wasted when the WHZ was ,<-3 SD. The data
for the children with extreme anthropometric values [e.g.,
Authors conducted the present study to determine the WHZ and height-for-age (HA) z-score values, <-6 and
performance of MUAC compared with WHZ for >.6] were excluded from the analysis. Descriptive
predicting deaths among hospitalized children aged 1 statistics were generated for all measurements and
year to 5 years in Indian settings. Authors also aimed to indexes. Sensitivity, specificity, and positive and negative
determine the best cut-off value of MUAC to predict predictive values for MUAC were generated for the WHZ
mortality in these children. at -3 and -2 cutoffs. Receiver operative characteristic
curves were generated for MUAC and WHZ.
METHODS
RESULTS
This study included primary data analysis of
anthropometric datasets from five nutrition surveys The total number of children included in the study was
conducted in four Indian states between 2016 and 2018. 1466 of whom 45.2% were girls. The mean±SD age was
Cross‑sectional study was conducted in Pediatric ward, 19.8±13.6 mo, WHZ was -1.21±1.37, weight-for-age
Civil Hospital, Ahmedabad. This nutrition study was (WA) z score was -1.87±1.39, HA z score was -

International Journal of Contemporary Pediatrics | March-April 2019 | Vol 6 | Issue 2 Page 523
 Mehta N et al. Int J Contemp Pediatr. 2019 Mar;6(2):522-526

1.61±1.46, and MUAC was 132±13mm (Table 1). There weight-for-height z score; WLZ, weight-for-length z
was a significant correlation between MUAC and WHZ score.
(Pearson correlation: 0.517, P<0.001). The area under the
receiver operating curve (AUC) was used to identify the For children in the age group from 6-24 mo, the mean
most appropriate choice for cutoffs that related MUAC (95% CI) AUC for MUAC was 0.841 (0.817, 0.862) at a
with the WHZ by selecting the point of highest WHZ of -3 and 0.829 (0.821, 0.838) at a WHZ of -2
cumulative value for sensitivity and specificity. Age- (P<0.001). The most appropriate MUAC cutoffs were,
stratified analyses in the 2 age groups (6-24 and 25-60 <120mm for a WHZ <-3 and <125mm for a WHZ <-2
mo) revealed the following results. with a sensitivity of 68.3% and 64.7%, respectively, and
a specificity of 86.3% and 83.4%, respectively. When the
Table 1: Anthropometric characteristics (N=1,466). MUAC cutoffs in current use were applied, the
sensitivities were 49.2% and 61.6%, respectively (Table
Minimum Maximum Mean±SD 2).
Age (mo) 6 60 19.8±13.6
Length (cm) 46.2 118.7 76.7±9.98 For children aged 25-60 mo, the mean (95% CI) AUC for
Weight (kg) 3.17 24.2 9.04±2.31 MUAC was 0.874 (0.841, 0.898) at a WHZ of <-3 and
WHZ -5.93 5.96 -1.21±1.37 0.853 (0.832, 0.868) at a WHZ of <-2 (P<0.001). The
WAZ -6.58 5.12 1.87±1.39 most appropriate MUAC cutoffs were <135mm for a
HAZ -5.98 5.87 -1.61±1.46 WHZ, <-3 and <140mm for a WHZ <-2 with a sensitivity
MUAC (mm) 71 254 132±13 of 63.7% and 65.4%, respectively, and a specificity of
81.6% and 78.3%, respectively. When the MUAC cutoffs
HAZ, height-for-age z score; MUAC, mid upper arm in current use were applied, the sensitivities were 23.15%
circumference; WAZ, weight-for-age z score; WHZ, and 28.7%, respectively (Table 3).

Table 2: Evaluation of screening test of nutritional status by different cutoffs of MUAC and WHZ (to detect severe
and moderate wasting status) in children aged 6-24 mo.

Positive Negative
Ages 6–24 mo* Sensitivity Specificity
prediction value predictive value
MUAC <115 mm for WHZ <-3 (current practice) 49.2% 91.7% 38.2 94.6
MUAC <120 mm for WHZ <-3 (proposed) 68.3% 86.3% 24.7 97.5
MUAC <125 mm for WHZ <-2 (current practice) 61.6% 82.6% 57.5 88.4
MUAC <125 mm for WHZ <-2 (proposed) 64.7% 83.4% 58.3 87.2
*Values in parentheses are exact MUAC cutoffs with the highest sensitivity and specificity at different WHZ cutoffs in different age
groups. To make it easier to remember, the rounded values for MUAC cutoffs are suggested.

Table 3: Evaluation of screening test of nutritional status by different cutoffs of MUAC and WHZ (to detect severe
and moderate wasting status) in children aged 25-60 mo.

Positive Negative
Ages 25–60 mo* Sensitivity Specificity
prediction value predictive value
MUAC <115 mm for WHZ <-3 (current practice) 23.1% 98.7% 52.1 96.9
MUAC <135 mm for WHZ <-3 (proposed) 63.7% 81.6% 14.9 98.8
MUAC <125 mm for WHZ <-2 (current practice) 28.7% 99.1% 80.2 84.4
MUAC <140 mm for WHZ <-2 (proposed) 65.4% 78.3% 43.9 92.6
*Values in parentheses are exact MUAC cutoffs with the highest sensitivity and specificity at different WHZ cutoffs in different age
groups. To make it easier to remember, the rounded values for MUAC cutoffs are suggested.

DISCUSSION the best comparison between the 2 methods. With the use
of the WHZ as the comparator, authors showed that
Large samples of 2 groups of children to carry out a current guidelines for viewing with MUAC failed to
statistical comparison of the WHZ and MUAC to identify a considerable proportion of children at risk and
categorize children at risk of SAM and MAM. The that the magnitude of the mismatch increased with age.
samples were drawn from the hospital setting and When the cutoff of <115mm recommended by the WHO
stratified by age. The AUC for receiver operative was used, for children aged between 6 and 24 mo the
distinctiveness was used to recognize the values that gave sensitivity for the detection of SAM was 49.2%; for

International Journal of Contemporary Pediatrics | March-April 2019 | Vol 6 | Issue 2 Page 524
 Mehta N et al. Int J Contemp Pediatr. 2019 Mar;6(2):522-526

children aged and for children aged 25–60 mo the MUACAZ, HA.11 They also founded that, under field
sensitivity was 23.1%. conditions, minimally trained workers made fewer and
smaller errors with MUAC than with WA or WH. 11,12 An
For children aged 6-24 mo, authors identified a cutoff for important operational improvement for the use of MUAC
MUAC of <120mm to categorize SAM. Of the 835 is that the same cutoff is used for all children. even
children in this age range, 41 subjects had severe wasting though MUAC increases with age and height, correcting
on the basis of a WHZ <-3, but only 32 children would arm circumference for either of the variables MUAC-for-
have been identified with the use of an MUAC cutoff age or MUAC-for-height does not offer any advantage as
<115mm compared with a total of 142 children who a predictive indicator for mortality.13,14 For this reason,
would have been identified had the cutoff been <120mm. community-based programs generally use a single
Similar reflection would apply for recommended WHO MUAC admission threshold without adjustment for age.
MUAC cutoff <125mm been used for the detection of MUAC-based programs may also recruit younger
children with MAM. children, but this may be beneficial because these
children are also the most vulnerable to illness and at
For the 631 children aged 25-60 mo, 34 subjects had higher risk of death.15
severe wasting on the basis of a WHZ <-3, and only 13 of
these children would have been identified if an MUAC It has been reported that the prevalence of SAM on the
cutoff <115mm was sed. A cutoff <135mm would have basis of a WHZ <-3 or MUAC <115mm is similar, but
detected 72 children. With the use of the recommended the comparisons were not based on measurements that
<125-mm cutoff for the detection of MAM, 221 children were made in the same children.14 Both MUAC and the
had moderate or severe wasting on the basis of a WHZ <- WHZ are used as alternative for more complex changes
2 but only 23 of these children would have been in aspects of body composition, but they do not capture
identified if the MUAC cutoff <125mm was used the same changes. Therefore, differences might be
compared with 291 children if the cutoff <140mm was expected. Differences in the WHZ might be accounted for
used. by differences in leg length that are not directly related to
wasting and are not representative of important body
The findings are clear that, with the use of the currently compositional changes for a malnourished child, thus
recommended WHO cutoff for MUAC, a significant making it less reliable in the identification of children at
number of children would not have been identified as high risk.16,17 MUAC captures aspects of muscle mass
either severely or moderately malnourished compared and fat mass, and a close relation between MUAC and
with the number if the WHZ cutoff was used. MUAC has muscle mass has been indicated on the basis of an
clear benefits for the screening of nutritional status in assessment via dual-energy X-ray absorptiometry.
large numbers of children and in community-based Evidence in support of the use of MUAC as a single
programs such as growth monitoring and promotion indicator has come from a study in Kenya where it was
activities that are carried out by frontline health workers. shown that the use of both the WHZ and MUAC did not
improve the detection of high-risk undernourished
A comparison of the data for the hospital group with children.18 More recently, Briend et al suggested that
those of the community group indicated a lower MUAC there is no programmatic benefit in using both MUAC
cutoff for WHZs <-3 and <-2 in the hospital group (<123 <115mm or a WHZ <-3 to identify high-risk children.8 If
and <128mm, respectively) than in the community group a higher sensitivity is required for programmatic reasons
(<131 and <135mm, respectively). For both groups, the (i.e., to take into account poor food security), it seems
sensitivity and specificity were similar to those in the preferable to increase the MUAC cutoff rather than to
combined analysis. The higher cutoff that was resulting combine it with the WHZ. In the same way, if a higher
from the community group was attributed to the older age specificity is required, in case of inadequate treatment
of the children register in the community. 10 Because, both capacity, lowering the MUAC cutoff should be
in the hospital and community, the programs for preferable. The use of color-banded MUAC straps could
screening and management of children suffering from also help to minimize measurement errors.15 Experience
SAM have used a single age group of 6-59 mo, authors from Burkina Faso has reported that, as an admission
suggest the use of the results of MUAC cutoffs from the criterion for SAM, the use of a cutoff of 118mm for
combined data for hospital and community children. For MUAC was a useful alternative to the WHZ.19
many years, MUAC has been used as an alternative
indicator of nutritional status and has shown great utility CONCLUSION
in challenging situations such as during emergencies,
famines, or refugee crisis. Velzeboer et al, in 1983, tested The analysis reported here raises concerns that the use of
the dependability (i.e., precision) of 5 minimally trained the currently recommended cutoffs for MUAC in
community health volunteers in rural Guatemala for identifying children aged between 6 and 60 mo who are
weight-for-height (WH), HA, WA, MUAC, and suffering from SAM or MAM (<115 and <125mm,
midupper arm circumference for age z score (MUACAZ) respectively) are unlikely to identify many of the children
and reported that, under field conditions, intraobserver who would benefit from therapeutic care and, thus,
reliability was highest for WA followed by MUAC, remain vulnerable to the complications associated with

International Journal of Contemporary Pediatrics | March-April 2019 | Vol 6 | Issue 2 Page 525
 Mehta N et al. Int J Contemp Pediatr. 2019 Mar;6(2):522-526

malnutrition and are at increased risk of death. A revised, 8. Briend A, Maire B, Fontaine O, Garenne M. Mid-
higher value for the cutoff for MUAC in 2 age groups (6- upper arm circumference and weight-for-height to
24 and 25-60 mo) would include these vulnerable identify high-risk malnourished under-five children.
children On the basis of these observations, the present Matern Child Nutr. 2012;8:130-3.
study indicates that cutoffs for MUAC (i.e.<120mm for 9. Grellety E, Golden MH. Weight-for-height and mid-
SAM and <125mm for MAM in the 6-24 mo age group; upper-arm circumference should be used
and <125mm for SAM and <135mm for MAM in the 25- independently to diagnose acute malnutrition: policy
60 mo age group) would better capture vulnerability and implications. BMC Nutr. 2016;2(1):10.
10. National Nutrition Programme. Baseline survey 2004
risk. The use of these cutoffs would be a simple, reliable
report. Dhaka (Bangladesh): ICDDR, B. Available at:
alternative to the WHZ in identifying children suffering
http://www.popline.org/node/265812.
from severe wasting (SAM) and moderate wasting 11. Velzeboer MI, Selwyn BJ, Sargent F, Pollitt E,
(MAM). An obvious strength of the present study is that Delgado H. The use of arm circumference in simplified
the data came from a adequate sample. However, it screening for acute malnutrition by minimally trained
would be worth exploring the outcomes, including health workers. J Trop Pediatr. 1983;29:159-66.
morbidity and mortality or subsequent growth, 12. Briend A, Zimicki S. Validation of arm circumference
development, and body composition, with the use of the as an indicator of risk of death in one to four year old
MUAC cutoff compared with the existing gold standard children. Nutr Res. 1986;6:249-61.
in the children aged 5 y. The findings from this analysis 13. Briend A, Garenne M, Maire B, Fontaine O, Dieng K.
would provide a more secure basis for policy formulation Nutritional status, age and survival: the muscle mass
and advice for practice. hypothesis. Eur J Clin Nutr. 1989;43:715-26.
14. Pelletier DL, Low JW, Johnson FC, Msukwa LA.
Funding: No funding sources Child anthropometry and mortality in Malawi: testing
Conflict of interest: None declared for effect modification by age and length of follow-up
Ethical approval: The study was approved by the and confounding by socioeconomic factors. J Nutr.
Institutional Ethics Committee 1994;124:2082S-105S.
15. Myatt M, Khara T, Collins S. A review of methods to
REFERENCES detect cases of severely malnourished children in the
community for their admission into community-based
1. The Child Health Epidemiology Reference Group: therapeutic care programs. Food Nutr Bull.
Definitive global childhood causes of death estimates 2006;27:S7-23.
released by CHERG. Not on track to meet Millennium 16. Myatt M, Duffield A, Seal A, Pasteur F. The effect of
Development Goal 4. 2012. Available at: body shape on weight-for-height and mid-upper arm
http://www.jhsph.edu/departments/ international- circumference based case definitions of acute
health/the globe/archive/summer2012/childcauses-of- malnutrition in Ethiopian children. Ann Hum Biol.
death.html. 2009;36:5-20.
2. Behera SK, Mohapatra SS, Kar S, Das D, Panda C. 17. Brambilla P, Rolland-Cachera MF, Testolin C, Briend
Incidence and mortality of hospitalized diarrhea cases. A, Salvatoni A, Testolin G, et al. Lean mass of
Part III. Indian Pediatr. 1980;17:607-12. children in various nutritional states. Comparison
3. Sehgal V, Sethi GR, Sachdev HP, Satyanarayana L. between dual-energy X-ray absorptiometry and
Predictors of mortality in subjects hospitalized with anthropometry. Ann N Y Acad Sci. 2000;904:433-6.
acute lower respiratory tract infections. Indian Pediatr. 18. Collins S, Sadler K, Dent N, Khara T, Guerrero S,
1997;34:213-9. Myatt M, et al. Key issues in the success of
4. United Nations Development Programme :(2006) community-based management of severe malnutrition.
About MDGs: what they are. Available at: Food Nutr Bull. 2006;27:S49-82.
http://www.unmillenniumproject.org/goals/index.htm. 19. Goossens S, Bekele Y, Yun O, Harczi G, Ouannes M,
5. World Health Organization: WHO Child Growth Shepherd S. Midupper arm circumference based
Standards and the Identification of Severe Acute nutrition programming: evidence for a new approach in
Malnutrition in Infants and Children: A Joint regions with high burden of acute malnutrition. PLoS
Statement by the World Health Organization and the One. 2012;7:e49320.
United Nations Children’s Fund. Geneva: WHO. 2009
6. Fernández MÁ, Delchevalerie P, Van Herp M.
Accuracy of MUAC in the detection of severe wasting
with the new WHO growth standards. Pediatrics. Cite this article as: Mehta N, Bhatt RG, Vora H,
2010;126:e195-201. Parmar B. Comparison of midupper arm
7. Bern C, Zucker JR, Perkins BA, Otieno J, Oloo AJ, circumference and weight-for-height z score for
Yip R. Assessment of potential indicators for protein assessing acute malnutrition in children aged 6-60
energy malnutrition in the algorithm for integrated months: an analytical study. Int J Contemp Pediatr
management of childhood illness. Bull World Health 2019;6:522-6.
Organ. 1997;75(Suppl1):87-96.

International Journal of Contemporary Pediatrics | March-April 2019 | Vol 6 | Issue 2 Page 526