|BRIEF RESEARCH ARTICLE
|Year : 2016 | Volume
| Issue : 1 | Page : 68-72
Exploring the multidimensional nature of anthropometric indicators for under-five children in India
Ashish Kumar Gupta, Kakoli Borkotoky
Doctoral Candidate, International Institute for Population Science (IIPS), Mumbai, Maharashtra, India
|Date of Web Publication||23-Feb-2016|
Ashish Kumar Gupta
Doctoral Candidate, International Institute for Population Science (IIPS), Mumbai, Maharashtra
Source of Support: None, Conflict of Interest: None
| Abstract|| |
This study examined the multidimensional nature of the association of stunting, wasting, and underweight for children below 5 years of age in India using data from the National Family Health Survey (NFHS)-3 (2005-2006). Multiple correspondence analysis (MCA) was applied to examine the association of the indicators. Additionally, log-linear model was used to find out the model of best fit to examine the nutritional status of children. It was found that underweight is associated with both stunting and wasting, whereas there was no consistent pattern of association between stunting and wasting. The results also confirmed that children suffered from multiple anthropometric failures. The results showed that height-for-age, weight-for-height, and weight-for-age taken together give the model of best fit for analysis of nutritional status. The study concluded that the three indicators of nutritional status should be considered simultaneously to determine the percentage of undernourished children.
Keywords: Anthropometric failure, anthropometric indicator, log-linear model, multiple correspondence analysis (MCA), nutritional status, stunting, underweight, wasting
|How to cite this article:|
Gupta AK, Borkotoky K. Exploring the multidimensional nature of anthropometric indicators for under-five children in India. Indian J Public Health 2016;60:68-72
|How to cite this URL:|
Gupta AK, Borkotoky K. Exploring the multidimensional nature of anthropometric indicators for under-five children in India. Indian J Public Health [serial online] 2016 [cited 2020 Nov 25];60:68-72. Available from: https://www.ijph.in/text.asp?2016/60/1/68/177319
The nutritional status of children is an important indicator of health and development. In developing countries, malnutrition is a widespread problem, being an underlying cause of more than 2.2 million child deaths and one-fifth of disability-adjusted life years lost.  Therefore, one of the major objectives of the Millennium Development Goals is to reduce the proportion of malnourished children through the reduction in poverty and hunger. In India, out of the total children below 5 years of age, almost half are stunted, one out of every five is wasted, and 43% are underweight.  Earlier studies also reported that India has the largest number of undernourished children (63 million) in the world. , According to a recent United Nations Children's Fund (UNICEF) estimate, India accounts for 31% of the developing world's children who are stunted and 42% of those who are underweight.
Stunting (low height-for-age), wasting (low weight-for-height), and underweight (low weight-for-age) are the three most commonly used indicators to examine the nutritional status of children. Stunting indicates long-term nutritional deprivation; it does not depend much on recent food shortages or shifts in economic conditions. ,, Wasting is an indicator for acute malnutrition and underweight is a composite index of chronic or acute malnutrition. , While much has been written about the causes and effects of stunting and wasting in general, the relationship between wasting and stunting remain poorly understood. Little is known about the pathways, which lead to a child experiencing one or more of the conditions of anthropometric failure. Although height and weight are used to create anthropometric index, the three anthropometric indicators have their own limitations.  The use of weight-for-age as an indicator of nutritional status is recommended only in populations where accurate age information is available. 
It is seen that children who are stunted are more prone to be wasted and underweight. This is possible that once a child is stunted, it is difficult to revise it in late childhood.  Children identified with more than one form of anthropometric failure may have significant impact on the growth and development of the individual and the country at large. In view of the wide use of the anthropometric indicators to assess the nutritional status of children, this study focused on the multidimensional nature of association of the anthropometric indicators. Better understanding of the nature of association among the indicators will help in developing focused interventions to improve child health and survival.
The present study is based on the Indian National Family Health Survey (NFHS)-3 data conducted in 2005-2006. The survey was coordinated by the International Institute for Population Sciences (IIPS) and Macro International under the stewardship of the Ministry of Health and Family Welfare (MoHFW), India. The urban and rural samples within each state were drawn separately following a multistage sampling design. In each state, the rural sample was selected in two stages and urban sample in three stages. NFHS-3 collected information on height and weight of 51,555 children below 5 years of age. Additional information on sample design, including sampling framework and sample implementation, are given elsewhere. 
Children below 6 months of age were excluded from this study, assuming that till the age of 6 months breastfeeding is the main source of nutrition and they will be less affected by other socioeconomic conditions of the household. The anthropometric indicators were compared with the World Health Organization (WHO) Reference Population (2006). Children whose Z-score was below -2 standard deviation (SD) from the median of the reference population were considered as undernourished (stunted, wasted, and underweight).
The study first applied multiple correspondence analysis (MCA) to examine the nature of association among the three anthropometric indicators. Later, log-linear analysis was used to examine the model of best fit to examine malnutrition from a 2 Χ 2 Χ 2 contingency table. To select a log-linear model that adequately describes the structure of the association between underweight, stunting and wasting, the study applied stepwise model selection procedures. Chi-square and G 2 statistics were calculated to examine the model fit. Additionally, Bayesian information criterion (BIC) was used as model inclusion indicator. Lower the value of BIC, better the model in explaining pattern of association. In the first step, unsaturated models were examined with all possible two-way and three-way interactions and finally, the saturated model was used to examine all possible interactions at the same time. In addition, the coexistence of anthropometric failure was examined for children from different socioeconomic backgrounds to support the findings on the multidimensional nature of anthropometric indicators. Further, the states were classified into six regions following the criteria used in NFHS-3.
The two-dimensional display of the six categories of stunting, wasting, and underweight obtained through MCA are given in [Figure 1]. The figure showed that stunting and wasting lie in the second and fourth quadrants and their counterparts not-stunting, not-wasting, and not-underweight were in the first and third quadrants. On the other hand, underweight lay at the border line of the second and fourth quadrants. Positioning of the points indicated that there was no association of stunting and wasting while underweight was associated with both stunting and wasting.
|Figure 1: Multidimensionality of the indicators of nutritional status of children, India|
Click here to view
The study also observed that in India, children with different socioeconomic backgrounds suffered from more than one form of anthropometric failure [Table 1]. The results indicated that out of the total undernourished children, one-fourth was identified with both stunting and underweight and 9% have failed in all the three indicators. The results also showed that 24% boys and 27% girls were both stunted and underweight. Another 10% boys and 8% girls were identified as stunted, wasted, and underweight simultaneously. It was seen that the percentage of children identified as both stunted and underweight increased with an increase in age. On the other hand, the percentage of children being both wasted and underweight declined with an increase in age. The percentage of children with failure in all three indicators increased during 12-23 months of age and then declined. It was also seen that the percentage of children with multiple anthropometric failures was high among the poor, children with less educated mothers, and children from households that did not use improved drinking water, cooking fuel, and improved toilet facilities. Households that did not use improved source of cooking fuel and improved toilet facilities had more than two times higher percentage of children with multiple anthropometric failures. The highest and lowest percentages of children who were simultaneously stunted and underweight were seen in the central and southern regions, respectively. The central region had the second highest percentage of children who were stunted, wasted, and underweight at the same time followed by the eastern region.
|Table 1: Percentage of children (6-59 months) with more than one anthropometric failure in India|
Click here to view
The likelihood ratios chi-square (G 2 ) and BIC from nine log-linear models for checking the association of nutrition indicators are reported in [Table 2]. Model 1 is a baseline model, which assumed the marginal distribution of stunting, wasting, and underweight. Model 1showed the complete independence structure of components of the nutritional status with zero interaction terms. The G 2 and BIC indicated that the model did not fit well. Model 2 provided the partial independence structure of nutritional status for children under 5 years. It included joint independence of undernutrition with interaction of stunting and wasting. The values of G 2 and BIC indicated that the model does not adequately fit, i.e., underweight is partially independent of the combined effect of stunting and wasting. Similarly, Model 3 considered interaction terms of stunting and underweight with a marginal distribution of wasting. Moreover, G 2 and BIC indicated that wasting was independent of stunting and undernutrition. Model 4 indicated that stunting was partially independent of wasting and undernutrition. Models 5-7 checked the conditional independence of one factor with other two factors. The G 2 and BIC in Model 7 showed that, the goodness of fit increased than other conditional independence models. Model 8 considered the interaction of each factor with other factors without an interaction term of all three factors. The values of G 2 and BIC indicated better fit of the model than the above discussed models. The saturated model (Model 9) contained the three factors being analyzed and all possible interactions between these factors. On the basis of BIC and G 2 values, Model 9 was identified as the best fit among all the models discussed in the study. 
|Table 2: Goodness-of-fi ts tests for log-linear models relating stunting, wasting, and underweight, India|
Click here to view
The results of the study confirmed the multidimensional nature of the anthropometric indicators of children. Underweight was associated with both stunting and wasting but stunting and wasting were not interrelated. This indicated that stunting was not necessarily a predictor for wasting but for those children who were stunted in their early childhood, it would be difficult to improve in late childhood. The findings on the association of stunting, wasting, and underweight were consistent with the findings from past research. , The results also suggested that the three-factor interaction model of stunting, wasting, and underweight was the best fitting model to study the nutritional status of children. Thus, these three factors should be considered simultaneously to get an accurate representation of the nutritional status of children. Additionally, multiple anthropometric failures could occur to children with different socioeconomic backgrounds. Thus, the study concludes that the selection of indicators to examine the nutritional status of children should be specific to the goal of the study because no single indicator can give comprehensive information of the nutritional status of children.
One of the limitations of the study is that the data is cross-sectional in nature, which limits the ability to establish a causal relationship. Hence, it was possible that the short-term variations in weight due to some illness may affect the estimates. Additionally, some information on height was missing and some were out of the acceptable range; so we had to exclude all those information.
Financial support and sponsorship
The authors had no financial support to conduct the research.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, et al
. Maternal and child undernutrition: Global and regional exposures and health consequences. Lancet 2008;371:243-60.
Arnold F, Parasuraman S, Arokiasamy P, Kothari M. Nutrition in India. National Family Health Survey (NFHS-3), India, 2005-06. Mumbai, India: International Institute for Population Sciences; 2009. p. 59.
FAO, WFP, IFAD. The State of Food Insecurity in the World 2012. Economic Growth is Necessary but Not Sufficient to Accelerate Reduction of Hunger and Malnutrition. Rome, FAO; 2012. Available from: http://www.fao.org/docrep/016/i3027e/i3027e.pdf. [Last accessed on 2014 May 20].
Ministry of Statistics and Programme Implementation (MOSPI). Children in India 2012 - A Statistical Appraisal 2012. Available from: http://mospi.nic.in/mospi_new/upload/Children_in_India_2012.pdf. [Last accessed on 2014 May 23].
United Nation's Children Fund (UNICEF). Improving Child Nutrition: The Achievable Imperative for Global Progress 2013. Available from: https://washnutrition.files.wordpress.com/2013/05/unicef-nutrition_report_final_2013.pdfon. [Last accessed on 2014 Jun 06].
Ramachandran P, Gopalan HS. Assessment of nutritional status in Indian preschool children using WHO 2006 Growth Standards. Indian J Med Res 2011;134:47-53.
World Health Organization (WHO). Nutrition Landscape Information System (NLIS): Country Profile Indicators, Interpretation Guide 2010. Available from: http://whqlibdoc.who.int/publications/2010/9789241599955_eng.pdf. [Last accessed on 2014 Jun 06].
Gorstein J, Sullivan K, Yip R, de Onís M, Trowbridge F, Fajans P, et al
. Issues in the assessment of nutritional status using anthropometry. Bull World Health Organ 1994;72:273-83.
Wolde M, Berhan Y, Chala A. Determinants of underweight, stunting and wasting among schoolchildren. BMC Public Health 2015;15:8.
Nguefack TG, Léonie DN. Multidimensional nature of undernutrition: A statistical approach. J Med Med Sci 2011;2:690-5.
[Table 1], [Table 2]
|This article has been cited by|
||Nutritional Status of Infants 0-23 Months of Age and its Relationship with Socioeconomic Factors in Pangkep
| ||Veni Hadju,Rabina Yunus,Andi Imam Arundhana,Andi Ummu Salmah,Atjo Wahyu |
| ||Asian Journal of Clinical Nutrition. 2017; 9(2): 71 |
|[Pubmed] | [DOI]|
||Policy content and stakeholder network analysis for infant and young child feeding in India
| ||Seema Puri,Sylvia Fernandez,Amrita Puranik,Deepika Anand,Abhay Gaidhane,Zahiruddin Quazi Syed,Archana Patel,Shahadat Uddin,Anne Marie Thow |
| ||BMC Public Health. 2017; 17(S2) |
|[Pubmed] | [DOI]|