Malnutrition: The silent crisis: Collation of anthropometric measures of nutritional status in children under 2 years of age: A hospital-based case–control study

ABSTRACT

Background:

Even if malnourishment is a life-threatening condition, there is not a single ‘gold standard’ anthropometric measurement to diagnose child undernutrition. Hence, this case–control study compared different anthropometric measurements to assess child malnutrition.

Methodology:

Using WHO’s MGRS Criteria 2006, cases and controls were selected and matching was done for age and sex. The calculated sample size was 154 (77 cases and 77 controls), assuming a two-sided confidence level of 95%, power of the study 80% and a case–control ratio of 1:1, 10% nonresponse rate and lack of exclusive breastfeeding taken as an exposure factor. The Z-scores (WFH, HFA, WFA) were calculated using WHO Anthro software. The sensitivity, specificity, and accuracy were calculated for each anthropometric measure. Multiple linear regressions for comparison of MUAC against WFA and HFA Z-scores were performed. The Composite Index of Anthropometric Failure (CIAF) was also calculated.

Results:

Even among controls, 26% were severely stunted and 14.2% were severely underweight. The sensitivity MUAC to diagnose severely underweight and severely stunted children was 84.2% and 58.5%, respectively. Multiple Linear regression found positive association of WAZ and MUAC. Out of 154 children, 114 (74%) had anthropometric failure.

Conclusions:

The combined anthropometric measurements approach to screen chronic malnutrition in the community is strongly recommended. There is a need to develop software in the local language that is simple and feasible to use by grass route workers for early diagnosis of SAM children.

Introduction

Malnourishment is a life-threatening condition and acute malnutrition is an indicator of an emergency that requires urgent action.[1] According to UNICEF, there are 144 million stunted and 37 million wasted children under the age of five in the world.[2] In India, as per National Family Health Survey-5 (NFHS-5) prevalence of wasting, stunting and underweight is 19.3%, 35.5%, and 32.1%, respectively.[3] The children in their first 2 years are more vulnerable to undernutrition than toddlers or preschool children due to the high demand for nutrients, poor child-feeding practices, and infection,[2] so child growth faltering is high between the age of 0 to 24 months.[4]

The nutritional status of a child can be assessed by different methods. The most recent and updated indicators by the World Health Organization (WHO) Multicentric Growth Reference Study (MGRS) Criteria 2006[5] are in use which provides only individual assessment of stunting, wasting, and underweight. However, this approach fails to identify children who qualify for multiple categories of malnutrition.[5,6,7]

Mid-Upper Arm Circumference (MUAC) is also used by field workers for early diagnosis of malnutrition. However, a growing body of research indicates that MGRS Criteria and MUAC identify very different subgroups of malnourished children, with little overlap. Recently, it has been recognized that there is not a single ‘gold standard’ anthropometric definition of undernutrition; all definitions have pros and cons.

For primary healthcare provider, it is difficult to decide on a single method to estimate child malnutrition at a population level. Hence, this study compared different anthropometric measurements for assessing malnutrition with a hospital-based case–control study

Methodology

A hospital-based case–control study was conducted from April 2021 to April 2022 at Civil Hospital, Rajkot, Gujarat. The permission form the Institutional Ethical Committee (IEC) had been taken before study conduction.

By using WHO’s MGRS Criteria 2006,[5] children who had weight for height Z-score (WHZ) <−3 SD with or without nutritional oedema and were admitted to the Nutritional Rehabilitation Centre (NRC) were included as cases. Controls were healthy children attending the Immunization clinic with WHZ between −2 SD and +1 SD or MUAC ≥ 13.5 cm. Only those children whose mothers had given consent and who were free from any chronic illness and congenital anomalies were included in the study. Cases and controls ratio 1: 1 was maintained as both groups were matched for age and sex.

Lack of exclusive breastfeeding was taken as an exposure factor where a proportion of cases (SAM diagnosed child) not exclusively breastfed was 44.4% with an odds ratio (OR) of 2.8 from a case–control study conducted at Vellore, Southern India.[6] The calculated sample size was 154 (77 cases and 77 controls) by using the STATCAL application of Epi-Info, assuming a two-sided confidence level at 95%, power of the study 80%, and a case–control ratio of 1:1 and 10% nonresponse rate.[7]

Mothers of all study participants were interviewed for sociodemographic details and feeding practices of child. The weight and height of children were measured by an infantometer. MUAC was measured by Shakir’s tape. Data entry and analysis were done in Microsoft Office Excel 2019. Multiple Linear regression, for comparison of MUAC against WFA and HFA Z-scores, was performed using the software package Epi Info (Version 7.2.2.6) from CDC, Atlanta, USA.

Z-score was calculated for each child by using WHO Anthro software (version 3.2.2, 2011).[8] It is an online tool developed by WHO, which helped to calculate Height for Age (HFA), Weight for Age (WFA), and WFH Z-scores of each child. The children were also classified in 7 groups based on Composite Index of Anthropometric Failure (CIAF), developed by Svedberg and modified by Nandy et al.[9] Three indices proposed by Bose and Mandal[10] were also used to assess the problem of stunting, underweight, and wasting relative to the total prevalence of undernutrition, Stunting Index (SI = Stunting/CIAF), Wasting Index (WI = Wasting/CIAF) and Underweight Index (UI = Underweight/CIAF).

Result

In this study, most study participants (79.2%) were of 7–24 months of age group. Among cases and controls, 53.3% were male and 46.7% were female. Education of parents found that parents of cases (40.3% mothers and 31.2% fathers) were more illiterate. Among parents’ occupation of cases, only 68.8% were homemakers and 71.4% of cases had a father working as a labourer. Most study participants (62.3% cases and 70.1% controls) had received breastfeeding as soon as after birth; 87% cases and 92.2% controls had received colostrum feeding. Among cases between 7 months and 2 years of age, only 47.6% were given semisolid or solid diet as complementary feeding (CF), as compared with controls (65.6%) [Table 1].

Table 1.

Sociodemographic and feeding details of study participants

Variables	Cases (n=77), n (%)	Controls (n=77), n (%)
Mother’s education
Illiterate	31 (40.3)	6 (7.8)
Literate	46 (59.7)	71 (92.2)
Father’s education
Illiterate	24 (31.2)	3 (3.9)
Literate	53 (68.8)	74 (96.1)
Mother’s occupation
Homemaker	53 (68.8)	74 (96.1)
Labourer/Service	24 (31.2)	3 (3.9)
Father’s occupation
Labourer	55 (71.4)	29 (37.7)
Service	22 (28.6)	48 (62.3)
Pre-lacteal feeding
Not received	64 (83.1)	73 (94.8)
Received	13 (16.9)	4 (5.2)
Time of initiation of BF
As soon as after delivery	48 (62.3)	54 (70.1)
After some times	29 (37.7)	23 (29.9)
Colostrum feeding
Received	67 (87.0)	71 (92.2)
Not received	10 (13.0)	6 (7.8)
Variables*	Cases (n=61), n (%)	Controls (n=61), n (%)
Age of starting of complementary feeding
At 6 months	16 (26.3)	21 (34.4)
After 6 months	45 (73.7)	40 (65.6)
Type of food given as complementary feeding
Solid/Semi solid	29 (47.6)	40 (65.6)
Liquid	32 (52.4)	21 (34.4)

*Participants between 7 months to 2 years of age are 61 in number

Among cases, 31 (40.2%) were severely stunted and 62 (80.5%) were severely underweight, while among controls, 20 (26%) were severely stunted and 11 (14.2%) were severely underweight [Table 2].

Table 2.

Distribution of study participants according to HAZ and WAZ scores

	Cases (n=77), n (%)	Controls (n=77), n (%)
HAZ*
<−3SD (severely stunted)	31 (40.2)	20 (26)
−2 to −3 SD (stunted)	15 (19.6)	12 (15.6)
−2 to +2 SD (normal HAZ)	31 (40.2)	45 (58.4)
WAZ*
<−3SD (severely underweight)	62 (80.5)	11 (14.2)
−2 to −3 SD (underweight)	9 (11.7)	6 (7.8)
−2 to 0 SD (normal WAZ)	6 (7.8)	60 (78)

*HAZ=Hight for age Z-score, WAZ=Weight for age Z-score

The sensitivity and specificity of MUAC to diagnose severely underweight children between 7 months and 2 years were 84.2% and 80%, respectively. However, the sensitivity and specificity of MUAC to diagnose severely stunted children between 7 months and 2 years were low (only 58.5% and 53.8%, respectively) [Table 3].

Table 3.

Comparison of MUAC against WAZ and HAZ scores (7 months to 2 years of age)*

MUAC	WAZ		Total
	<−3SD	>−3SD
<11.5	48	13	61
>11.5	9	52	61
Total	57	65	122*
Sensitivity: 84.21%, Specificity: 80%, PPV: 78.7%, NPV: 85.3%, Accuracy: 82%
MUAC	HAZ		Total
	<−3SD	>−3SD
<11.5	24	42	61
>11.5	17	49	61
Total	31	91	122*

Sensitivity: 58.5%, Specificity: 53.8%, PPV: 36.4%, NPV: 74.2%, Accuracy: 55.3%. *Total was 122 as MUAC was taken for children between 7 months to 2 years of age

In this study, multiple linear regression was performed for comparison of MUAC against WFA and HFA Z-scores for children between 7 months and 2 years of age. In both cases and controls, WAZ was positively significant with a P value < 0.05. While HAZ was negatively significant with P value < 0.05, which indicates that MUAC alone is not adequate to screen stunting in children [Table 4].

Table 4.

Multiple linear regression for MUAC with WAZ and HAZ scores of cases and controls (7 months to 2 years of age)*

	Coefficient	F-test	P
Anthropometric indicators of cases
HAZ	−0.3	11.7	0.00
WAZ	0.8	24.9	0.00
Anthropometric indicators of controls
HAZ	−0.1	4.1	0.04
WAZ	0.3	7.4	0.00

*MUAC was taken for children between 7 months to 2 years of age

According to CIAF classification of the 7 groups, among cases, only 6 (7.8%) had purely wasting, while among controls, 4 (5.2%) had only wasting, 3 (3.9%) had both wasting and underweight, 14 (18.1%) had stunting with underweight and 16 (20.8%) had only stunting. In the final CIAF, all 77 cases (100%) and 37 controls (48%) had anthropometric failure. So out of 154 children, 114 (74%) had anthropometric failure. From CIAF, out of 154 study participants, SI was 0.14 and WI was 0.71 [Table 5].

Table 5.

Classification of children according to CIAF* with CIAF indices

Group name	Cases (n=77), n (%)	Controls (n=77), n (%)	Total (n=154), n (%)
A (No failure)	0 (0)	40 (52)	40 (26)
B (Wasting only)	6 (7.8)	4 (5.2)	10 (6.5)
C (Wasting with underweight)	25 (32.5)	3 (3.9)	28 (18.2)
D (Wasting, stunting with underweight)	46 (59.7)	0 (0)	46 (29.9)
E (Stunting with underweight)	0 (0)	14 (18.1)	14 (9)
F (Stunting only)	0 (0)	16 (20.8)	16 (10.4)
Y (Underweight only)	0 (0)	0 (0)	0 (0)
CIAF (sum of group B to Y)*	77 (100)	37 (48.1)	114 (74)
CIAF index	Cases	Controls	Total
Stunting index (SI) = Stunting/CIAF	0/77=0	16/37=0.43	16/114=0.14
Underweight index (UI) = Underweight/CIAF	0/77=0	0/37=0	0/114=0
Wasting index (WI) = Wasting/CIAF	77/77=1	4/37=0.11	81/114=0.71

*Composite Index of Anthropometric Failure (CIAF), CIAF derived by summation of groups B to Y

Discussion

Among controls, 15.6% were stunted, 26% severely stunted, 14.2% were severely underweight and 7.8% were underweight. It indicates that even if the MUAC of a child is normal, a child may suffer from stunting or wasting in a hidden form. Extensive literature searching, a multicountry study conducted by Grellety and Golden[11] and a study published in BMC from 47 different countries showed that the number of children diagnosed by one criterion or the other varied dramatically across countries. The variation in the prevalence of acute malnutrition based on these two indicators (MUAC and WHZ) could, therefore, be associated with different aspects such as body composition.[12]

In this study, the sensitivity to diagnose severely underweight and severely stunted children between 7 months and 2 years was 84.2% and only 58.5%, respectively. A cross-sectional survey conducted by Kalawati Saran Children Hospital, in collaboration with UNICEF India, a study conducted at GMERS Medical College and General Hospital, Vadodara, Gujarat,[13] and a study done at Sir T. Hospital, Bhavnagar, Gujarat,[14] also found that a number of children diagnosed with only MUAC were less as compared with children diagnosed with only WHZ or by both MUAC and WHZ.

In the study, multiple linear regression found that WFA Z-scores of cases and controls were positively associated with MUAC (P value < 0.05). HFA Z-scores were negatively associated (P value < 0.05) with MUAC (coefficient for cases −0.29 and for controls −0.14). It suggests that the use of MUAC alone is less effective for diagnosing stunting in children. MUAC can be helpful in diagnosing wasting and underweight; however, its effectiveness in diagnosing stunting is limited. The presence of either normal or abnormal MUAC findings may lead us to conclude that a child is healthy, despite the possibility that the child is experiencing chronic malnutrition.

In this study, overall CIAF (out of 154 children) found that 74% of children had anthropometric failure. Out of this 74% CIAF, most cases and controls fall in Group D, which means all three: stunting, wasting and undernutrition were present. The high anthropometric failure was seen in a study conducted by Dr. Goswami[15] at Odisha, India, (52.3% CIAF) and also 29.6% CIAF seen among children of 0 to 24 months of age residing at Bogor District of Indonesia.[16] The study conducted in Chhattisgarh, India, also reported high anthropometric failure (62.1%) among children between 12 and 35 months of age.[17] In Ethiopia[18] 48.5%, in India[19] 49.1% and in Bangladesh[19] 39% anthropometric failure had been observed among children of <5 years of age. The higher prevalence of CIAF attributed to the fact that children in their first 2 years are more vulnerable to undernutrition than toddlers or preschool.[2] The conventional methods used to assess malnutrition in a population underestimate the prevalence of malnutrition compared with CIAF.[17] This method also identifies children with single or multiple anthropometric failures.[20]

In the study, Stunting Index (SI), Wasting Index (WI) and Underweight Index (UI) were calculated from CIAF. The findings revealed 0.14 SI, 0.71 WI and 0 UI. Notably, there was not a single child suffering from underweight solely, and it was found that children were suffering from underweight with stunting or wasting. The studies conducted by Nandy et al., Seetharaman et al., Bose et al.[10] and Gupta et al.[20] also found similar indices. The underweight indicates both acute as well as chronic malnutrition; however, it does not provide whether the condition is exclusively acute or chronic.[14] The utility of these three indices is that they inform us about the relative severity of stunting, underweight and wasting with respect to total undernutrition in a population. It suggests that the higher the value, the greater the severity concerning total undernutrition. By excluding normal individuals, these indices are undernutrition-specific.[10] It must be pointed out that these indices cannot replace the conventional measures of undernutrition. Rather they should supplement them to get a more comprehensive picture of the nutritional stress being experienced by a population.[15]

This study compares the WHO criteria, the MUAC and the CIAF index. Among these three, the MUAC is currently used by field workers; however, it reflects only recent episodes of child malnutrition and does not provide information about overlapping malnutrition categories. The CIAF index is quite difficult to count for the grass root workers. Therefore, the WHO criteria serve most effective tool for malnutrition detection. In cases where resources are adequate, a combined approach may be implemented.

Conclusion

The combined anthropometric measurements approach especially to screen chronic malnutrition in the community is strongly recommended instead of relying exclusively on one anthropometric measurement and it will also increase the power of diagnosis. It will help in early diagnosis and prompt treatment of malnutrition which in turn will reduce child morbidity and mortality. By using WHO Anthro software as a reference, similar software in the local language can be made for grass route workers to help in the early diagnosis of malnutrition by grass route workers and primary health care providers.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.