Concordance between weight-for-height z-score (WHZ) and mid-upper arm circumference (MUAC) for the detection of wasting among children in Bangladesh host communities
By Md. Lalan Miah, Dr. Md. Khalilur Rahman, Dr. Md. Abdul Alim and Bijoy Sarker
Md. Lalan Miah is the Nutrition Surveillance Manager for Action Against Hunger Bangladesh where he has seven years’ experience leading the nutrition surveillance project in both host communities and refugee camps. Md. Lalan also leads the Nutrition Sector Assessment Technical Working Group (ATWG).
Dr. Md. Khalilur Rahman is Director General of Bangladesh National Nutrition Council (BNNC) where he supports the development of national nutrition policy and programmes with multiple government and non-governmental partners. He is a medical doctor with over 25 years of work experience in the government healthcare system.
Dr. Md. Abdul Alim is an Assistant Professor and Deputy Program Manager in National Nutrition Services, Institute of Public Health Nutrition (IPHN), Dhaka where he has worked in public health and nutrition since 2010.
Bijoy Sarker is a Public Health Nutritionist currently working as SMART Regional Advisor for Asia with Action Against Hunger Canada. Bijoy has nearly a decade of experience working in emergency nutrition and surveillance in multiple countries across Africa and South Asia.
The authors would like to acknowledge the following individuals and agencies for their technical support and collaboration on this project including National Nutrition Services (NNS), Institute of Public Health Nutrition (IPHN), Ministry of Health and Family Welfare (MoHFW) Bangladesh, Bangladesh National Nutrition Council (BNNC), Action Against Hunger Bangladesh Mission (Jogie Abucejo Agbogan, Fitsum Tesfaye, Dr. Md Abul Hasan, Md. Saiful Islam Talukder and G.M. Mosharaf Hossain), Action Against Hunger France HQ (Clemence Malet and Benjamin Guesdon), Technical Rapid Response Team (Alexandra Humphreys) and Global SMART Initiative and Action Against Hunger Canada (Jana Daher and Hassan Ali Ahmed).
What we know: There is poor concordance between children identified as wasted using weight-for-height z-score (WHZ) and mid-upper arm circumference (MUAC). MUAC only programming is recommended in Bangladesh community-based management guidelines.
What this article adds: A secondary data analysis explored the concordance between WHZ and MUAC to identify wasting in 12 districts of Bangladesh and the impact of different MUAC thresholds to capture low WHZ children. Analysis found poor and varied concordance between the two indicators with WHZ identifying the highest proportion of wasted children. Greater MUAC thresholds increased the proportions of severely and moderately wasted cases identified but also led to greater ‘false positives’ (children not meeting WHZ nor MUAC criteria). Expanded MUAC criteria ≥115 mm misclassified a proportion of severe acute malnutrition (SAM) cases (WHZ <-3) as eligible for moderate treatment. This has implications for the capacity and resources of existing programmes and reinforces questions regarding outcomes of low WHZ children excluded or under-treated by MUAC only programming.
Both weight for height z-score (WHZ) and mid-upper arm circumference (MUAC) are recommended by the World Health Organization (WHO) as independent indicators of wasting in children (WHO 2009; WHO 2013). In practice, MUAC is increasingly used as a stand-alone anthropometric criterion,1 such as for community-based screenings, at remote clinics where height/length boards and weighing scales may not be available and as a temporary programming adaptation to COVID-19 where Infection Protection Control (IPC) measures cannot be assured (UNICEF, WHO 2020). MUAC and WHZ often identify different children, with the degree of overlap varying greatly by context (Grellety et al, 2016). While MUAC has long been considered as a better indicator of risk of death than WHZ (Briend et al, 2016), recent analyses suggest that children with low WHZ and low MUAC have a similar risk of dying (Schwinger et al, 2019; Grellety and Golden, 2018a and 2018b). Thus, there remain questions regarding the outcomes of low WHZ children that are not identified in MUAC only strategies, especially in contexts where concordance between WHZ and MUAC is poor and how to manage this, such as increasing the MUAC threshold to capture low WHZ children.
In Bangladesh, there are two national guidelines for the management of wasted children: i) facility-based management of children with severe acute malnutrition (SAM)2 and ii) community-based management of acute malnutrition (CMAM). The facility-based SAM guideline recommends the admission and treatment of wasted children by WHZ/weight-for-length (WLZ) and/or MUAC as per WHO guidelines (WHZ/WLZ <-3SD and/or MUAC <115 mm and/or bilateral pitting oedema including medical complications) (IPHN, MOHFW, 2017). The national CMAM guideline, covering the management of uncomplicated SAM, moderate acute malnutrition (MAM) and acutely malnourished pregnant and lactating women, recommends admission and treatment using the MUAC only criterion (IPHN, MOHFW, 2017).
Given the above, Action Against Hunger (ACF) undertook an in-depth analysis of anthropometric data from multiple surveys across different regions of Bangladesh to investigate concordance between WHZ and MUAC in the detection of wasting to help decision-making with regards to CMAM programming in the Bangladesh national context.
After the recent publication of a similar analysis based on a large set of surveys across multiple countries, including Bangladesh (Guesdon et al, 2020),3 the main objective of this study was to explore the concordance of wasting between WHZ and MUAC in Bangladeshi children to inform policy and programming. In addition, the analysis explored the effect of increasing MUAC referral and admission cut-offs on the concordance between low MUAC and low WHZ/WLZ and the implications for the number of expected admissions to outpatient therapeutic feeding programmes (OTP) (severe wasting without medical complications) or to targeted supplementary feeding programmes (TSFP) (moderate wasting).
Figure 1: Profile of SMART surveys by district in Bangladesh
Data were drawn from population representative SMART surveys on the nutritional status of Bangladeshi children aged 6-59 months conducted by Action Against Hunger (ACF) Bangladesh between 2009 and 2020. A total of 34 SMART surveys across 12 districts of Bangladesh (Figure 1) rated either “excellent” or “good” quality and with an acceptable standard deviation for WHZ (standard deviation between 0.80 and 1.20)3 were retained for analysis. Two SMART surveys with the Emergency Nutrition Assessment (ENA) plausibility score “acceptable” were excluded. Individual child observations missing MUAC, WHZ, or oedema data (n=342), children with WHZ with outliers based on World Health Organization (WHO) flags (±5 SD, n=10), extreme MUAC (n=3) and children with bilateral pitting oedema (n=1) were excluded from the analysis.4 In total, the pooled sample for analysis consisted of 16,519 children 6-59 months.
A limitation of the study was that data was drawn from a limited number of districts, surveys were often small-scale and infrequently representative of the entire district and were conducted during different seasons of the year. Furthermore, the final sample was not weighted, so larger samples had a stronger influence on the pooled results.
The final analyses included 16,519 children aged 6 to 59 months of which 2,349 cases of wasting (1,996 MAM and 353 SAM) were identified by WHZ and/or MUAC.5 Table 1 shows the concordance of wasting prevalence by WHZ only, by MUAC only and by both WHZ and MUAC.
According to the all-district pooled analysis (Figures 2 and 3), it was estimated that overall 63.5% of GAM and 65.4% of SAM cases were identified by WHZ only; 15.1% of GAM and 19.3% of SAM cases were identified by MUAC only and 21.4% of GAM and 15.3% of SAM cases were identified by both WHZ and MUAC.
Figure 2: Concordance between WHZ and MUAC for GAM case detection
Figure 3: Concordance between WHZ and MUAC for SAM case detection
Table 1: Number of surveys and children included by district and concordance of wasting case detection (GAM and SAM) by WHZ only, MUAConly and both WHZ and MUAC
* This data is not representative of the entire district.
Among all surveyed districts (Table 1), the highest concordance for GAM was observed in Habiganj (27.0%) and Cox’s Bazar (25.6%). The lowest concordance for GAM was observed in Sirajganj (6.7%) and Satkhira (8.4%) districts. The highest concordance for SAM was observed in Sunamganj (20.0%) and Cox’s Bazar (19.8%). In three district samples, all SAM cases were identified by either WHZ or MUAC (no cases identified by both). In three district samples, all SAM cases were identified by WHZ alone. These findings indicate poor and varied concordance between the two indicators with WHZ identifying the highest proportion of wasted children.
Data was further analysed to project different scenarios of SAM and MAM case detection (Figures 4 and 5) and admission eligibility (Figure 6) for OTP and TSFP programmes based on different MUAC cut-offs. Figure 4 shows that, based on the global reference criteria for SAM (WHZ <-3 SD and/or MUAC <115 mm) and MAM (-3 SD ≤ WHZ < -2 SD /or 115 mm ≤ MUAC < 125 mm ), a standalone MUAC only referral cut-off of <115mm would be able to detect 34.6% of total SAM (by both MUAC and WHZ). A standalone MUAC only referral cut-off of <125 mm would be able to identify only 32.1% of total MAM (by both MUAC and WHZ) of the all-district pooled sample. Increasing the standalone MUAC only referral cut-offs further increases both SAM and MAM case detection (by both MUAC and WHZ); for example, a MUAC of <140 mm would capture 83.7% of MAM cases (by MUAC and/or WHZ criteria for MAM).
Figure 4: Projection of SAM case detection based different MUAC referral cut-offs
“SAM ALL Detected’’ refers to all cases of SAM by WHZ (<-3 SD) and/or MUAC (<115 mm). . “SAM ALL Undetected" refers to those children who are SAM (WHZ <-3 SD) by WHZ with a MUAC above the cut-off used for screening.
Figure 5: Projection of MAM case detection based different MUAC referral cut-offs
“MAM ALL Detected’’ refers to all cases of MAM by WHZ (-3 SD ≤ WHZ < -2 SD) and/or MUAC (115 mm ≤ MUAC < 125 mm) and not SAM by any indicator. “MAM ALL Undetected’’ refers to those children who are r MAM (-3 SD ≤WHZ < -2 SD) by WHZ with a MUAC above the cut-off used for screening.
As the aim of the expanded thresholds is to try to capture children with low WHZ not detected using standard MUAC cut-offs, we defined cases that met the increased MUAC cut-off criteria but did not meet the WHZ criteria as ‘false positives’. There are practical consequences of false positive cases. Children may be referred from the community based on MUAC screening but considered nourished in a nutrition facility if re-assessed using WHZ and returned home. This may undermine the screening strategy. In the absence of WHZ verification at facility-level, the nutrition status of MUAC referred children would not be verified and hence admitted for treatment. This has wide implications for programme capacity and resources.
Figure 6 shows the proportion of children eligible for SAM and MAM programmes including false positives based on different referral cut-offs. Among all referred cases with an expanded MUAC cut-off of <120 mm, 55% children met admission eligibility for the SAM programme and 45% for the MAM programme as per the global thresholds for MUAC and WHZ. However, the analysis estimates that increasing MUAC referral cut-offs from <125 mm to <130 mm, for example, would lead to 36% false positives (those with a MUAC ≥ 125 mm and WHZ ≥ -2 SD) who would not meet admission eligibility for the SAM and MAM programmes. The percentage of false positives would increase to 70% with a raised MUAC referral cut-off of <140 mm (Figure 6).
Figure 6: Proportion of children eligible for SAM and MAM programmes including false positives from different MUAC referral cut-offs (based on both MUAC and WHZ treatment eligibility as per global protocol)
*False Positive cases are those children whose MUAC ≥125mm and WHZ ≥ -2 based on global threshold for wasting.
The expanded MUAC only admission criteria were also applied to different scenarios to understand the programmatic implications (Figures 7 and 8). The implication of the expanded MUAC only programming is the misclassification of a number of MAM cases as eligible for OTP admission. Scenario 2 in Figure 7 shows that an expanded MUAC only admission criteria with a raised cut-off of <120 mm will lead to the admission of 45% of MAM children to OTP (115 mm ≤ MUAC < 120 mm and/or -3 SD ≤ WHZ < -2 SD). A raised MUAC cut-off of <12 5mm for OTP admission will lead to the OTP caseload consisting of 75% MAM children (those with 115 mm ≤ MUAC < 125 mm and/or -3 SD ≤ WHZ < -2 SD).
Figure 7: Proportion of SAM and MAM children within the OTP programme based on different expanded MUAC only programming criteria
For TSFP, the expanded MUAC only admission also has programmatic implications with the misclassification of a number of SAM cases as eligible for TSFP (those with a WHZ<-3SD and a MUAC ≥115 mm), and with a large proportion of false positives included when the admission cut-off exceeds 125 mm (Figure 8). This will further inflate the caseload, as many well-nourished children will be enrolled into the TSFP programme due to the expanded MUAC admission criteria.
Figure 8: Pie charts presenting the proportion of SAM, MAM and non-acutely malnourished children within the TSFP programme target, based on different expanded MUAC only programming
Discussion and Conclusion
Discordance between wasting prevalence as identified by WHZ and MUAC among children 6-59 months is well documented in the literature (Grellety, 2016 and Bilukha, 2018). This is confirmed by the present analysis in 12 districts of Bangladesh where the majority of children were wasted by WHZ only compared to MUAC only or by both criteria. Using only one indicator of wasting (MUAC only or WHZ only) for the diagnosis and treatment of wasting will result in a significant proportion of malnourished children being undetected and excluded from treatment. For Bangladesh, this analysis indicates that more than 60% of wasted cases (wasted by WHZ only: GAM 63.5%; SAM 65.4%) would not be detected using MUAC-only diagnostic criterion in CMAM programme. In other words, more than three of every five cases of wasting were excluded when WHZ was not considered as a diagnostic criterion.
In countries where screening and admission is conducted exclusively using MUAC and the concordance of wasting between WHZ and MUAC is poor, the overall cut-off for referral has sometimes been raised in an effort to capture more GAM by WHZ cases. This study demonstrates that the higher a MUAC cut-off is raised, the greater the number of wasting cases by WHZ will be detected. However, this has side effects of including non-wasted children and MAM children in SAM treatment programmes. These increased referrals have the potential to overload treatment centre capacity and accelerate opportunity cost in terms of availability of resources due to providing SAM treatment to MAM children.
The present study explored different scenarios to understand the allocation of SAM and MAM cases into OTP and TSFP programmes based on different expanded MUAC only programming where WHZ is not measured (due to national protocol) or not possible (as in COVID-19 in some contexts where Infection Protection Control (IPC) standards cannot be met). If an expanded MUAC threshold for screening is applied, the threshold must be carefully selected considering the proportion of misclassification and false positive cases introduced by expanded MUAC only cut-offs. This requires a thorough in-country consultation, capacity gap assessment and informed decision-making that takes into account the capacity of health and nutrition service centres (staff, supplies and other resources) to ensure that they are not overburdened. The absence of WHZ as a detection and verification method excludes a large proportion of wasted children from treatment or risks overloading treatment capacity with a high proportion of false positive cases if higher MUAC cut-offs are used. This reinforces questions regarding outcomes of low WHZ children excluded by MUAC only programming.
In Bangladesh, the national CMAM guideline recommends use of MUAC as the sole anthropometric criterion to identify SAM for admission to outpatient sites across the country. This may reflect that MUAC has practical advantages in terms of community screening and a lack of data on discordance between WHZ and MUAC when the national CMAM guideline was developed in 2017. Given this, we hope this analysis can help to provide updated information and understanding around the technical issues around the poor concordance between WHZ and MUAC in Bangladesh specifically.
For more information please contact Md. Lalan Miah.
1 Or presence of bilateral pitting oedema
2 The terms acute malnutrition and wasting are used interchangeably in this article in keeping with terminology used nationally.
3 as per the Emergency Nutrition Assessment (ENA) plausibility score
4 Due to the physiological effects of oedema causing retention of fluid in the body, weight data was not considered for children with nutritional oedema.
5 Prevalence could not be calculated given that this is a pooled analysis across different years.
Bilukha, O., Leidman, E. (2018). Concordance between the estimates of wasting measured by weight-for-height and by mid-upper arm circumference for classification of severity of nutrition crisis: analysis of population-representative surveys from humanitarian settings
Briend et al. (2016). Low mid-upper arm circumference identifies children with a high risk of death who should be the priority target for treatment. BMC Nutrition (2016) 2:63 DOI 10.1186/s40795-016-0101-7
Grellety, E., Golden, M.H. (2016).Weight-for-height and mid-upper-arm circumference should be used independently to diagnose acute malnutrition: policy implications. BMC Nutr 2, 10
Grellety, E., Golden, M.H. (2018a). Severely malnourished children with a low weight for-height have a higher mortality than those with a low mid-upper-arm circumference: I. Empirical data demonstrates Simpson’s paradox. Nutr J. 2018;17:79.
Grellety, E., Golden, M.H. (2018b). Severely malnourished children with a low weight for-height have similar mortality to those with a low mid-upper-arm circumference: II. Systematic literature review and meta-analysis. Nutr J. 2018;17:80
Guesdon, B., Couture, A., Pantchova, D., Bilukha, O. (2020). Potential consequences of expanded MUAC-only programs on targeting of acutely malnourished children and ready-to-use-therapeutic-food allocation: lessons from cross-sectional surveys. BMC Nutr. 2020;6:5. Published 2020 Feb 10. doi:10.1186/s40795-019-0328-1
IPHN, MoHFW (2017) National guidelines for the facility-based management of children with severe acute malnutrition in Bangladesh
IPHN, MoHFW (2017) National guidelines for Community-based Management of Acute Malnutrition in Bangladesh
Schwinger, C., Golden, M.H., Grellety, E.G., Roberfroid, D., Guesdon B. (2019). Severe acute malnutrition and mortality in children in the community: comparison of indicators in a multi-country pooled analysis. PLoS One.;14(8): e219745.
UNICEF (2020). Prevention, early detection and treatment of wasting in children 0-59 months through national health systems in the context of COVID-19. UNICEF, New York.
UNICEF & WHO (2020). Prevention, Early Detection and Treatment of Wasting in Children 0-59 Months through National Health Systems in the Context of COVID-19, United Nations Children’s Fund and World Health Organization, New York, 2020.
More like this
Can one use the same thresholds for SAM/MAM prevalence when using MUAC as the basis for measurement? For example can one equally use less than 15% GAM prevalence to classify...
Posted on behalf of East Africa project staff: Has anyone made any estimates on the implications for caseload, if admission criteria are changed from MUAC <11cm to MUAC...
View this article as a pdf Lisez cet article en français ici By Alexandra Humphreys, Bijoy Sarker, Baidar Bakht Habib, Anteneh Gebremichael Dobamo and Danka...
What is the most appropriete practical decision should one make when confronted with cases of borderline MUACs and Z-scores of 11.5cm and -2 SD respectively in nutrition...
en-net: WFH versus MUAC
I would like experts input in this regard. I wish Mark Myatt to be one of the respondent of my question. Much has been said about the discrepancy of MUAC and WFH in some...
en-net: MUAC cutoff to screen SAM
Looking at the WHO growth standard for MUAC, one can understand the change of MUAC with age. So my concern is, is it feasible to use the same MUAC cutoff (<11.5CM) for all...
FEX: Substandard discharge rules in current severe acute malnutrition management protocols: An overlooked source of ineffectiveness for programmes?
View this article as a pdf Lisez cet article en français ici By Benjamin Guesdon and Dominique Roberfroid Benjamin Guesdon is a nutrition and health research advisor...
Summary of research1 Location: Bangladesh What we know: MUAC and weight-for-height z score (WHZ) are common anthropometric criteria to diagnose severe acute malnutrition...
FEX: Simplifying the response to childhood malnutrition: MSF’s experience with MUAC-based (and oedema) programming
By Kevin P.Q. Phelan, Candelaria Lanusse, Saskia van der Kam, Pascale Delchevalerie, Nathalie Avril and Kerstin Hanson Kevin P.Q. Phelan was the Nutrition Working Group Leader...
There has been a discussion about the use of ONLY MUAC as an admission and discharge criteria (http://www.en-net.org/question/468.aspx). Although I understand the challenge in...
View this article as a pdf Lisez cet article en français ici By Tracy Dube, Mary Chelang'at Koech, Piyali Mustaphi, Sandra Harlass, Jessica Bourdaire and Karanveer...
Author Koert Ritmeijer, MSF Holland In Hlaing Thayer township, Yangon, Burma, ORWs were spending a considerable amount of time doing weight and height measurements on all...
FEX: Weight-for-height and mid-upper-arm circumference should be used independently to diagnose acute malnutrition: policy implications
Summary of research1 Location: Global What we know: Overlap between mid-upper-arm circumference (MUAC) and weight-for-height Z-scores (WHZ) when assessing acute malnutrition...
The CMAM Forum is currently mapping out all programmes where MUAC is used as the only admission criteria for SAM treatment, along with bilateral pitting oedema (as opposed to...
Given the available knowledge on the impact of MAM treatment vs nutrition prevention activities on the reduction of child undernutrition and mortality, and the huge investment...
I have started this thread in the hope that we can come together there to work out what we need to do to move forward with an informed debate on issues around MUAC-only...
We're considering monitoring MUAC on admission but are having some issues in how to interpret it. 1) What would be the MUAC cut-off of early vs. late admissions? We've...
FEX: Assessment of the PROBIT approach for estimating the prevalence of acute malnutrition from population surveys
Summary of research1 Location: N/A What we know already:Prevalence of GAM is normally estimated using two stage cluster sampled surveys using the SMART method. The PROBIT...
Summary of research1 The World Health Organisation (WHO) and UNICEF propose to use two independent criteria for diagnosing non-oedematous severe acute malnutrition (SAM) in...
View this article as a pdf Lisez cet article en français ici By Odei Obeng Amoako Adobea Gloria, Franck Alé, Paul Binns, Kevin Phelan, Jose Luis Álvarez...
Reference this page
Md. Lalan Miah, Dr. Md. Khalilur Rahman, Dr. Md. Abdul Alim and Bijoy Sarker (). Concordance between weight-for-height z-score (WHZ) and mid-upper arm circumference (MUAC) for the detection of wasting among children in Bangladesh host communities. Field Exchange 63, October 2020. p81. www.ennonline.net/fex/63/whzmuacbangladesh