This article is mentioned in:

Introduction

Nutrition plays a crucial role in the growth of healthy children and is even more critical for pediatric patients to help increase response to treatment, reduce morbidity and mortality, improve quality of life, and reduce treatment costs. Malnutrition in inpatients is relatively common due to a number of influencing factors, such as the inflammatory response, reduced nutrient intake, increased metabolic requirements, malabsorption and psychosocial issues. These conditions may be related to acute conditions (trauma, burns, infections) or chronic conditions (cancer, chronic kidney disease, heart failure, inflammatory bowel disease, neurological disease (1).

Anthropometric indexes are an integral part of screening and assessment of the nutritional status (2). The use of appropriate tools, accurate measurement techniques and appropriate reference data are necessary for the collection and interpretation of anthropometric indexes. However, there are a number of challenges associated with the measurement of anthropometric indexes in inpatients due to disease severity, associated technical interventions such as respiratory support, hemodynamics, or clinical situations, such as trauma, burns and casts. Therefore, anthropometric indexes, such as weight, or height are sometimes based solely on a doctor's estimation or information provided by parents/caregivers, and may be affected by the fluid status of the body. By contrast, mid-upper arm circumference (MUAC) is a simple, low-cost and objective method of assessing nutritional status that has been widely used in the community to detect malnutrition. A recent consensus statement by the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition (AND/ASPEN) on the indicators recommended for the detection of childhood undernutrition admits that ‘MUAC is the predicted mortality with a higher sensitivity than the weight for height in malnourished pediatric patients’ (3,4). Nonetheless, the detection of malnutrition in the hospital by anthropometric indexes is still limited; thus, MUAC can be considered as an index for the screening and monitoring of the nutritional status. Therefore, the present study was conducted with the aim of investigating the correlation between the MUAC z-score and the anthropometric indexes, weight and height, in assessing the nutritional status of pediatric inpatients.

Patients and methods

Subjects

The present study was conducted on 500 pediatric inpatients from 2 to 60 months of age who were treated at the Vietnam National Hospital of Pediatrics (Hanoi, Vietnam) between August, 2021 and August, 2022. The exclusion criteria include patients with physical restrictions that precluded the determine of standing height or recumbent length. All guardians/parents of the children were explained about the purpose of the study and were required to sign the consent forms. Children information was kept completely confidential and used for research purposes only. The present study was conducted after the research protocol was approved by the Vietnam National Hospital of Pediatrics (Decision no. 646/BVNTW-HDDD).

Data collection

The nutritional status of the children was assessed through data collection using measurement methods based on the WHO 2006 guidelines. The nutritional status was then classified based on the z-score according to the WHO 2006 guidelines (5) as follows: i) Weight was measured using the UNICEF SECA floor scale to measure the weight of the pediatric inpatients (weighing accuracy, 100 g); ii) recumbent length was obtained in infants up to 24 months of age (6); iii) standing height was determined in children >24 months of age; iv) MUAC was measured using tape provided by the Clinical Nutrition Center of the Vietnam National Hospital of Pediatrics; this was used to measure the MUAC in cm and the z-score (MUAC accuracy, 1 mm). The midpoint of the arm was deemed as the midpoint of the segment from the sacral process to the superior process of the humerus; the midpoint was marked, and the tension of the ruler was then adjusted, not too tight or too loose. The color range corresponding to the age of the patient was determined. The MUAC and z-score results were then read. Anthropometric data (z-core results) were processed using WHO 2006 Anthro software (version 3.2, a web link for Anthro version 3.2 software: https://who-anthro.software.informer.com/3.2/#google_vignette). The MUAC measurements are illustrated in Fig. 1.

Figure 1 MUAC and z-score results. Orange: 2 to 3 MUAC z-score, moderately overweight; yellow: 1 to 2 MUAC z-score, slightly overweight; green: 0 to 1 MUAC z-score, normal; green + brick: -1 to 0 MUAC z-score, Normal; yellow + brick: -2 to -1 MUAC z-score, mild malnutrition; orange + brick: -3 to -2 MUAC z-score, moderate malnutrition; red + brick: -4 to -3 MUAC z-score, severe malnutrition. MUAC, mid-upper arm circumference.

Statistical analysis

Statistical analysis was performed using SPSS 20.0 software (IBM Corp.). Descriptive statistics of frequencies and percentages were used to describe qualitative variables. Quantitative variables are presented as the mean ± standard deviation. The Kappa coefficient and Pearson's correlation coefficient (r) were used to determine the correlation between the MUAC z-score and other anthropometric indexes. With the coefficient r, r#x003C;0.25 was considered to indicate a weak correlation, r=0.25-0.5 a moderate correlation, r=0.5-0.75 a strong correlation, and r≥0.75 to indicate a very strong correlation. With the Kappa coefficient, ≤0.2 was considered to indicate a slight consensus; 0.21-0.40 a fair consensus, 0.41-0.60 a moderate consensus, 0.61-0.80 a substantial consensus, and 0.81-1.00 to indicate a perfect consensus. Receiver operating characteristic (ROC) curve analysis for the MUAC z-score was performed based on weight/height ≥-2SD. Non-malnourished children were defined as having positive results. The area under the ROC curve (AUC) was calculated to compare the classification ability of the new index. The optimal MUAC z-score cut-off score was determined based on the highest sensitivity and specificity. An AUC from 0.5-0.6 was considered unsatisfactory, an AUC from 0.6-0.7 was considered satisfactory, an AUC from 0.7-0.8 was considered good, an AUC from 0.8-0.9 was considered very good, and an AUC of >0.9 was considered excellent. In all analytical results, a value of P#x003C;0.05 was considered to indicate a statistically significant difference.

Results

The general characteristics of the 500 pediatric inpatients participating in the study are presented in Table I. In terms of sex, the percentage of boys was greater than that of girls (at 59.6 and 40.4%, respectively). The median age of the study subjects was 17 months. The mean age of the children in the study group was 21.4±16.6 months. The age group of #x003C;12 months accounted for the highest rate, at 40.4% (Table I).

Table I General characteristics of the study subjects (n=500).

Table I

General characteristics of the study subjects (n=500).

General characteristics	No. of patients	%
Sex
Male	298	59.6
Female	202	40.4
Age group
1-#x003C;12 months	202	40.4
12-24 months	113	22.6
25-60 months	185	37
Mean ± SD	21,4±16.6
(minimum-maximum)	(2-59)

The nutritional status of the pediatric inpatients according to anthropometric indexes is presented in Table II. Of note, as regards weight for age, 20.8% of the patients were underweight, and 1.8% of the patients were overweight and obese. As regards weight for height, 14% of the patients were classified as wasting, and 5.4% of the patients were classified as overweight and obese. For height for age, 24% of the patients were classified as stunting; in 10.2% of these patients, this was considered as severe. As regards the MUAC z-score, 27.2% of the patients were classified as having malnutrition; in 14.4% of these patients, this was considered as severe malnutrition. Of note, 1% of the patients were classified as being overweight (Table II).

Table II Nutritional status of the pediatric inpatients according to anthropometric indexes (n=500).

Table II

Nutritional status of the pediatric inpatients according to anthropometric indexes (n=500).

	Weight for age		Height for age		Weight for height		MUAC
Nutritional status	No. of patients	%	No. of patients	%	No. of patients	%	No. of patients	%
Severe malnutrition	52	10.4	51	10.2	26	5.2	64	12.8
Moderate malnutrition	52	10.4	69	13.8	44	8.8	72	14.4
Normal	387	77.4	380	76	403	80.6	359	71.8
Overweight	7	1.4			18	3.6	5	1
Obese	2	0.4			9	1.8

[i] MUAC, mid-upper arm circumference.

The MUAC z-score was found to have a strong correlation with the weight/height z-score (r=0.608; P#x003C;0.001) (Fig. 2). The MUAC z-score was also found to have a moderate consensus with the WHO absolute value MUAC classification (k=0.458, P#x003C;0.001) (Table III).

Figure 2 MUAC z-score has a strong correlation with weight/height z-score (r=0.608; P#x003C;0.001). MUAC, mid-upper arm circumference.

Table III Consistency (consensus) between the MUAC z-score and the WHO MUAC cut-off (n=500).

Table III

Consistency (consensus) between the MUAC z-score and the WHO MUAC cut-off (n=500).

	MUAC z-score
	Severe malnutrition	Moderate malnutrition	Normal	Overweight	Total (n)	κ	P-value
MUAC						0.458	0.001
Severe malnutrition	39	6	0	0	45
Moderate malnutrition	12	13	14	0	39
Normal	13	53	345	5	416
Total (n)	64	72	359	5	500

[i] MUAC, mid-upper arm circumference.

Furthermore, ROC curve analysis with weight/height ≥-2SD is the standard index and children without malnutrition exhibited positive results; the optimal cut-off point for MUAC z-score was -1.88. The AUC was 0.84 (95% confidence interval), indicating that this cut-off has an 84% ability to distinguish between malnourished and non-malnourished children (Fig. 3).

Figure 3 ROC curve analysis for mid-upper arm circumference z-score based on weight/height ≥-2SD. The area under the ROC curve was 0.84 (95% confidence interval), indicating that this cut-off has an 84% ability to distinguish between malnourished and non-malnourished children. ROC, receiver operating characteristic.

Discussion

Nutritional status affects the response of each pediatric patient to disease. A poor nutritional status in inpatients is associated with the duration of treatment, morbidity and mortality. Therefore, nutritional assessment is an integral part of the overall treatment of all pediatric inpatients. To date, anthropometric indexes (weight, height and MUAC) have been considered essential and are the preferred methods used to assess the nutritional status and for the monitoring of nutritional interventions in the community. However, these anthropometric indexes when used in hospitals still have limitations due to disease status and the distribution of body fluids (1).

The choice of which instrument is appropriate in the hospital should be considered based on resources, implementation and accuracy. Therefore, in the present study, in order to provide data for the selection of anthropometric indexes in the early identification of the nutritional status of patients upon admission, from August, 2021 to August, 2022, a total of 500 pediatric inpatients at Vietnam National Hospital of Pediatrics participated in the study.

In terms of sex, the percentage of boys was higher than that of girls (at 59.6 and 40.4%, respectively). The mean age of the study group was 21±16 months old (the lowest was 2 months and the highest was 60 months) with a median age of 17 months (Table I). The group of children #x003C;12 months of age accounted for the highest percentage (40.4%), followed by the groups of 12-24 months (22.6%) and 25-60 months (37%). In the study by Huong et al (7) on 337 inpatients aged between 6 months and 15 years, the children #x003C;2 years of age accounted for the highest rate (59.3%), followed by those aged 24 to 59 months (22%), 5 to 9 years (13.1%) and >9 years (5.6%).

The results of the present study are based on the reference population of the WHO (2006). The rate of underweight malnutrition (weight/age) was 20.8%, that of chronic malnutrition (stunting) was 24% and that of acute prevalence (wasting) was 14% (Table II). Currently, malnutrition is one of the key child health concerns in Vietnam. According to data from the Ministry of Health in 2021(8), in the community, the rate of acute malnutrition (weight/height) in children was 5.2%, that of underweight malnutrition was 11.5%, and that of stunting (height/age) was 19.6%; this rate has decreased significantly compared to 2000 (8.6, 33.8 and 36.5%, respectively). However, the prevalence of malnutrition in the present study was higher than that reported in the community in the healthy group of children (14, 20.8 and 24%, respectively). The reason for this is that the causes of malnutrition in the community and hospitalized children are different. In the community, malnutrition is often caused by a lack of food resources, socio-economic and psychological factors, and a lack of medical care, so as the country develops, the rate of malnutrition in the community will increase. Malnutrition in pediatric inpatients due to nutritional imbalances includes decreased nutrient intake (fluid restriction, heart failure, anorexia), increased nutrient loss (chronic diarrhea toxicity, burns, proteinuria), or increased nutritional requirements (increased metabolism) (burns) disproportionate to intake, and failure to absorb or utilize nutrients (malabsorption phase).

In the MUAC classification (Table III), there were 27.2% malnourished children, of which 12.8% were classified as having severe acute malnutrition and 14.4% were moderately acute malnourished children. The study results revealed that the use of the MUAC z-score significantly increased the ability to detect acute malnutrition in inpatients compared with weight/height and weight/age indexes. By contrast, the study by Lan et al (9) revealed a lower prevalence of malnutrition according to arm circumference (10.8%); the reason for this was that Lan et al (9) used the MUAC cut-off value to assess malnutrition and this was not based on the z-score. Similarly, the study by Hop et al (10) also demonstrated that when using a cut-off point of MUAC to assess nutritional status, the sensitivity and specificity were only high in groups 6-12 months, while in the group of 37-60 months, the sensitivity decreased to only 4% in boys and 10% in girls. In addition, the study by Hossain et al (11) in Bangladesh also demonstrated that MUAC thresholds varied by age group and that the use of the MUAC z-score will address these differences when assessing undernutrition. Therefore, in 2014, the American Dietetic Association recommended the use of the MUAC z-score for the assessment of the nutritional status in pediatric clinical practice (12). When compared with the study using the MUAC z-score at the hospital, the research results were lower than those of the study by Linh et al (13) with the prevalence of malnutrition according to the arm circumference of 41.7%. The present study was on patients with both acute and chronic disease, while the study by Pham Thao Linh was only on chronic liver disease with a high percentage of muscle wasting.

The present study on 500 inpatients (Fig. 2) revealed a strong correlation between the MUAC z-score and weight/height z-score (r=0.608); this result was similar to that of the study by Roberfroid et al (14) with a strong correlation (r=0.638). The MUAC z-score is quite similar to the WHO absolute MUAC classification of nutritional status (with k=0.458 and P#x003C;0.001) (Table III). Previously, there were some studies demonstrating that there was no consensus. There is a clear agreement between anthropometric indices to determine nutritional status, with body mass index (BMI), weight/height and MUAC tending to determine the prevalence of undernutrition differently (14,15). This is likely due to the majority of studies comparing age/sex-specific criteria (weight/height z-score or BMI z-score) and absolute value-based criteria (MUAC in cm), and it has also been demonstrated that classification based on z-score has a higher sensitivity and less bias than the absolute cut-off (16). For example, in the study by Roberfroid et al (14), the sensitivity to detect malnourished children using MUAC 125 cm was 31%, and weight/height was 70.6%. The study by Laillou et al (16) demonstrated that when using the current WHO cut-off point of 115 mm to screen for malnutrition, >90% of children with weight/height #x003C;-3SD were missed.

In the present study, the AUC was 84.1%, indicating that this cut-off value was a good index of malnutrition. Based on the ROC curve analysis, the present study demonstrated that the MUACZ #x003C;-1.88 was a useful and simple nutritional index to assess malnutrition in children.

In conclusion, assessing nutritional status in inpatients is essential. It is necessary to use a combination of all three anthropometric indexes: Weight, height and MUAC to detect early clinical malnutrition. Particularly for patients whose weight is affected by fluid status, MUAC is an important tool for assessing the nutritional status.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

TMTL and LTT conceptualized the study. TMTL and HNV were involved in the study methodology. ATTV and HNV provided the software and were involved in the formal analysis. LTT was involved in the investigative aspects of the study. LTT and TMYL were involved in data collection. ATTV, HNV and LTT were involved in the writing and preparation of the original draft of the manuscript. HNV and LTT were involved in the writing, reviewing and editing of the manuscript. TMTL supervised the study. TMTL and HNV was involved in project administration. TMTL and LTT confirm the authenticity of all the raw data. All authors have read and agreed to the published version of the manuscript.

Ethics approval and consent to participate

All guardians/parents of the children were explained about the purpose of the study and were required to sign the consent forms. Children information was kept completely confidential and used for research purposes only. The present study was conducted after the research protocol was approved by the Vietnam National Hospital of Pediatrics (Decision no. 646/BVNTW-HDDD).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References