Do we need to reconsider the CMAM admission and discharge criteria?; an analysis of CMAM data in South Sudan

Forty nine million children of age 5 years or less were reported wasted in 2018 [1]. Though malnourishment among children aged 5 years or less has decreased by 10% since 1990, the prevalence of wasting was estimated at nearly 7.3% in 2018 [1]. Additionally, UNICEF reports that more than 1.5 million children mortality occurred due to severe acute malnutrition every year [2] and 3.5 million children of age 5 years or less with moderate acute malnutrition succumb to death in 2005 [3]. Acute malnutrition is defined by a decrease of two standard deviations (SD) below the WHZ (Weight for Height Z score) [4] while chronic malnutrition, described as stunting, is defined by a decrease of two SD below the HAZ (Height for Age Z score) [5].

For children aged between 6 and 59 months, the World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) have advised to use both WHZ and Mid Upper Arm Circumference (MUAC) as well as bipedal oedema for identification of children with Acute Malnutrition [6, 7]. However, MUAC tends to replace WHZ and to be considered as a single indicator in the community together with bipedal oedema for malnutrition screening [7, 8]. This trend is rationalized by MUAC’s superiority in predicting mortality and has other benefits such as simplicity and accuracy of the measurement, easy training and its attribute of being less affected by dehydration [6, 7, 9‐14]. Furthermore, the prognostic accuracy of MUAC to predict mortality is not inferior to that of combined screening by both MUAC and WHZ or even MUAC z-score for age [14, 15].

However, neither MUAC nor WHZ on its own screening is able to capture a significant portion of children who succumbed to malnutrition [14]. It is also recognized that each screening measure is poorly related [14, 16], and there still exists uncertainty about the characteristics of the two malnourished types by either MUAC or WHZ [17].

Furthermore, chronic irreversible sequelae including failure of intellectual development and stunting are not properly reflected in the current community screening guideline. There is some evidence showing that SAM-WHZ represents a potential stunting risk better than SAM-MUAC [18‐21], and several studies proved that if the children succeed in timely weight gain, following catch up growth could be observed after the treatment from nutrition program [22‐26]. In other words, unless appropriate management takes place in time, catch-up growth hardly occurs afterward [27]. If we use MUAC as the single modality for community screening of acute malnutrition, we might lose opportunity for catch-up growth for those who are excluded by MUAC criterion but still malnourished by WHZ criterion.

For discharge from Community Management of Acute Malnutrition (CMAM) program, WHO recommends to use the anthropometric criteria for admission [28]. Severely wasted children lose their muscle mass almost a half or two third of the amount expected for their height, which would indicate that muscle mass depletion is more often in the children with SAM [29, 30]. A few studies showed initial weight gain among the malnourished children who were hospitalized was primarily due to rapid fat mass increase rather than muscle mass gain [31‐33]. If these are also valid for the malnourished children enrolled in CMAM program, which is designed to identify early stage of malnourished cases and provide timely management in communities before developing complication to limit cases demanding hospitalization, simple recovery of anthropometric measures does not necessarily mean full restoration of each body composition. Furthermore, this would lead to avoidable recurrence and/or mortality after discharge from CMAM program considering that low muscle mass is the main mortality risk [33]. Several studies revealed unacceptable high mortality and recurrence rate after discharge from CMAM program [34‐37].

Here, this study aimed to characterize the clinical features of the malnourished children by diagnostic modalities; estimate correlation of each type of acute malnutrition with chronic malnutrition; and explore restoration pattern of body composition responding to management by analysing program and survey data.

As demonstrated in other series of studies [14, 16], our data also show poor correspondence between the two SAM’s (GAM’s) identified by MUAC and WHZ, displaying only 12.3% (26.5%) are overlapped. Similar to other studies [6, 17, 43], our study also shows that children with tall height are prone to be classified as malnourished by WHZ but normal by MUAC. This finding might suggest that some children with tall height are prone to be classified as malnourished by WHZ, while they do not have actual malnutrition. Even after exclusion of those children with SAM-WHZ with tall height, but normal by MUAC, the correlation estimation between each type of acute malnutrition and HAZ still showed higher between MUAC and HAZ than between WHZ and HAZ (0.185 vs. -0.096). This procedure prevents a biased decrease in correlation between acute malnutrition by WHZ and stunting, and leads to biased relative increase in correlation between MUAC and stunting. Besides our community survey data also demonstrate that MUAC identifies younger children and more girls as malnourished than WHZ, which is similar to observations in other studies [8, 44].

The community survey data show that the malnourished children by SAM-MUAC are more severely stunted (HAZ < − 3) than those by SAM-WHZ. Considering that the children with severe stunting would have the least chance to reach their genetic height even after the treatment, SAM-MUAC would not be a good screening modality to identify malnourished children for enrolment in a management program to prevent short stature. Here, we found that major portion of included children by low MUAC are already stunted and more loose criteria based on MUAC is required to capture a group of children with potential high risk in stunting. The observed correlation coefficient between MUAC and HAZ from community survey data is larger than that between WHZ and HAZ. Additional 2 × 2 contingency table analyses showed that there is a strong statistically significant association between SAM-MUAC and chronic malnutrition (HAZ < -3) (p = 1.274 × 10⁻⁷), but no statistically significant association was found between SAM-WHZ and chronic malnutrition (HAZ < -3) (p = 0.4576). This tendency persists even after exclusion of the tall children with SAM-WHZ but normal by MUAC. Together with the data from hospitalized children in Kenya showing that the children with SAM-MUAC tended to have a more chronic clinical feature (Kwashiorkor) than those by WHZ (5.0% vs. 49.0%, p < 0.001) [44], our finding supports that MUAC screening identifies more advanced malnourished cases than the screening by WHZ. Additionally, if children were with edema, there was strong geographical variations in their MUAC [45]. Thus, we might reconsider using the term of ‘acute’ for those malnourished cases identified by MUAC.

In early stage of wasting when a reasonable amount of fat in the body is retained, MUAC represents fat mass of the children [46]. In starvation, the body starts to use fat for energy source soon after the body faces insufficient energy intake [47]. However, body muscle mass is maintained at the later stage of malnutrition since protein catabolism is maintained at a minimum and the organism lives on fat body stores [48]. The median duration to progress to SAM is estimated about 7.5 month [49]. During this progression, the body mainly utilises fat for the energy source to support life. Therefore, in severely malnourished children with limited fat mass, the arm circumference represents remaining body muscle mass instead of fat [50]. Additionally, in severely malnourished children, infection is frequently super imposed [51‐55], when the reserved muscle mass is further to be utilized for amino acids resources for immune response as well as for energy resources [56].

Though WHZ screening identifies earlier malnutrition cases, it also has a limitation to identify normal children with tall height without malnutrition especially for the ethnic groups with long limbs as evidenced in this study.

The data also showed the relative velocity MUAC³ gain surpassed that of weight gain among the children with SAM both by WHZ and MUAC over the recovery period. Based on our data analysis and simple idea that fat has lower density than other body components, we suggest fat mass gain was faster than that of lean body mass during recovery period in South Sudan nutrition program. Also this conjecture is duplicated with the previous observation among severely malnourished children who were undergone parenteral nutrition for management [57]. Besides, considering that bone growth reflecting limb length does not occur simultaneously with weight and fat mass gain, MUAC³ gain could exaggerate real volume increase. However, the relative velocity of MUAC² gain consistently surpassed that of weight gain but in a lessor scale than MUAC³ gain (Fig. 4) with similar velocity peaks as in MUAC³ gain. Therefore, if mortality is related to decreased muscle mass [58], it is recommended to provide care for the malnourished until full muscle mass restoration even after the malnourished children has reached discharge criteria.

MUAC does not simply identify acute stage of malnutrition but also identifies more advanced malnourished cases than WHZ among the children in the South Sudan community. Our result supports that each screening modality by MUAC or WHZ identifies not only different characteristics but also different stages of malnutrition. Therefore, each screening modality should be used as a complementary to each other, as MUAC and WHZ identify different malnourished cases. Considering that the goal of community screening is to identify malnourished children in a timely manner and prevent not only mortality but also chronic complications including short stature, it is advised to employ both WHZ and MUAC measures at the community for acute malnutrition screening. Specifically, when MUAC is the only measure for the screening, it is recommended to use more loose MUAC cut-off value (133 mm) than current 125 mm to avoid excluding children who could be admitted by SAM-WHZ. This study also suggests initial weight gain is mainly ascribed to fat mass increase among the children with SAM, the current discharge criteria, restoration of anthropometric data still leave the malnourished children with insufficient restoration of muscle mass, and at risk of mortality after discharge. Given this, further research should be followed including assessment of body composition to revise the current admission and discharge criteria for CMAM program.