Does WHO child growth standards suitable for Indian Children: A comparison of anthropometric measurements from a cross-sectional survey

doi:10.21203/rs.3.rs-2738998/v1

Download PDF

Research Article

Does WHO child growth standards suitable for Indian Children: A comparison of anthropometric measurements from a cross-sectional survey

https://doi.org/10.21203/rs.3.rs-2738998/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:Growth charts are fundamental tools for monitoring individual infants and children. But the current WHO-specified growth standards generally rely on single, one-size-fits-all cut-offs.

Methods:The present study developed a new growth reference of Indian children aged 0-60 months using 2005-2006 & 2015-16 National Family and Health Survey data.

Results:Box-Cox power exponential (BCPE) distribution and lambda mu sigma (LMS) model suggested that the median height for Indian children at age five years was 6.35 cm (boys) and 4.23cm (girls) lower than the World Health Organization (WHO) growth reference values, 2006. A similar height value for age is observed in both Indian and WHO growth curves at younger generations. However, a significant height difference was observed in WHO growth curves compared to Indians at later ages. Relatively heavier children in the WHO standard appeared at later ages than the Indian growth curve.

Conclusions:The study recommended high priority in using national derived reference data, which correctly define growth trajectories in children.

Growth charts

NFHS

Box-Cox power exponential

HAZ score

WAZ score

India

Child nutritional status is an essential component of overall human development. Restricted growth in young children is a risk factor for mortality from infectious diseases and poor physical and cognitive development throughout life [1]. The percentage of childhood stunting has remained alarmingly high in many parts of India. In 2019-21, 35.5 percent of children below five years were stunted, and 32.1 percent were underweight, accounting for one in three children under five being stunted and underweight in India [2]. In India’s flagship nutrition program, POSHAN Abhiyaan, launched in 2017, stunting is one of the monitoring and evaluation metrics: underweight, low birth weight, and anemia [3]. The focus of government policies shifts to growth indicators rather than mortality indicators. By focusing on these indicators, governments can identify specific areas where intervention is needed to improve nutrition and prevent health problems before they occur.

Growth curves are among the most commonly used tools for assessing the general well-being of infants and children and the communities in which they live. In 2006, the World Health Organization (WHO) published a new worldwide growth standard statistical distribution for children aged 0 to 59 months [4]. But the WHO-specified relative child health measure stunted and underweight may need to be more accurate. It will not be significantly associated with absolute child health measures like infant and under-five mortality [5]. Height, weight, and BMI are inconsistent in human populations and within them [6]. Current standards for assessing both overnutrition and undernutrition have been based on single, one-size-fits-all cut-offs that disregard the impact of racial or ethnic disparities. Because of an inaccurate standard, a healthy child might be misclassified as malnourished, and he would be wrongly diagnosed and prescribed an increased diet which could prove harmful by turning the child obese. In developing nations, the WHO 2006 standards tend to over-diagnose stunting and wasting [7]. The main flaw of the current methodology is that it uses global height and weight levels to measure malnutrition, regardless of variations that might occur due to genetic, environmental, cultural, or geographic factors. Though medical literature recognizes the importance of these factors, the World Health Organization (WHO) ignores the effect of racial or ethnic differences. So, it is crucial to estimate the country or race-specific HAZ-score and WAZ-score, which will be consistent with the absolute measure of child health malnutrition. This study aimed to compare reference standards based on cross-sectional surveys for the growth of Indian children with World Health Organization (WHO) reference data.

Data Source

For the present analysis, anthropometric measures namely height/length (cm) and weight (kg) of children are used from two rounds of the National Family and Health Survey (NFHS 2005-2006 and NFHS 2015-16). NFHS provides information on a complete birth history, including the sex and date of birth of each child below age 5 years, from each ever-married woman in the sample from all the states. It also collected information on the height and weight of children under 5 years of age (under 3 years for NFHS-2) in all three survey rounds. The following criteria were used for the selection of the standard reference population:

Children lived in socioeconomic conditions favorable to growth (selecting the highest wealth score for top deciles which indicates the richest household).
Exclusive or predominant breastfeeding for at least 6 months.
No maternal smoking before and after delivery.
Single-term birth.

Of the total 2,80,889 children, 8,824 children aged 0-5 years were selected for the reference population.

Statistical Analysis

To generate standard Z-scores-stunting - low height for age (HAZ) and underweight - low weight for age (WAZ) for children aged 0 to 60 months, WHO (2006) used Box-Cox power exponential (BCPE) distribution. Hence, for the present study also, it is decided to use the same distribution and the final selected models simplified to the LMS (Lambda Mu Sigma) model. The LMS method was used to construct reference centile curves [8,9]. The curves were fitted as cubic splines by non-linear regression, using penalized likelihood; the extent of smoothing was controlled by equivalent degrees of freedom [10]. The required centiles (C) were calculated using the following equation,

Where L(t) is the Skewness (power of a Box-Cox transformation), M(t) is the median, S(t) is the coefficient of variation, and Z_α is the normal equivalent deviation corresponding to the centile. The z score of an individual measurement can be computed as

Further, growth curves will plot the comparison of Z-score and centile with WHO-based Z-scores for weight-for-age and height-for-age separately for boys and girls.

The estimated LMS parameters for height and weight by age and gender are presented in Table 1 and Table 2. The required centiles for each gender were calculated using these parameters. The predicted M values for weight and height in males were higher than the parameters in females for each month, as shown in Table 1 and Table 2, respectively. The updated growth reference values of the study indicated that boys are 1.5cm taller and 0.37 kg heavier than girls at 1 month of age. Boys are 0.46kg heavier and 1.76cm more elevated than their female counterparts at the age of one year. At, 59 months of age, boys are, on average, 0.23 cm shorter and 0.46 kg heavier than girls.

The height percentile curves derived from the LMS parameters are shown in Figures 1 and 2. Figures 3 and 4 show weight percentile curves produced from LMS parameters. The height and weight percentile curves for each gender show a steadily rising trend until 24 months. According to the analysis, it has been found that the estimated 50th percentile of height and weight values for boys and girls was significantly higher than the WHO percentile.

Figure 5, Figure 6, Figure 7, and Figure 8 compare height for age and weight-for-age Z-scores curves separately in India for boys and girls with the WHO 2006 reference Z scores. From the analysis, it has been shown that for Indian infants and children under five years, weight and length/height values were slightly higher than the WHO growth standards. The new construction showed lower values for each Z-score and percentile curve than WHO values. A significant height difference was observed in WHO growth curves compared to Indian growth curves at younger ages. However, at later ages (age 40 months onwards), similar height values for age are observed in both Indian and WHO growth curves. Relatively heavier children according to WHO standards have appeared at later ages as compared to the Indian growth curve. Growth patterns are similar for both boys and girls in India.

Table 3 shows the descriptive statistics for calculated Z scores and WHO Z scores for individual children in India. The mean estimated HAZ score was higher than the WHO HAZ score. Similarly, the mean calculated Z was higher than the WHO Z scores even in WAZ reported.

This study attempted to reformulate the Indian growth curve. It concludes that Indian growth standards significantly differ from WHO growth standards. This study's constructed growth chart values are closer to Khadilkar et al. (2007) [11]. The average height is higher for Indian children than the WHO growth standards, although these differences are disappearing at younger ages. It is also found that healthy Indian infants are heavier than the WHO norms. A similar observation was made by previous literature [11,12]. Using the WHO 2006 cut-offs, a higher percentage of boys and girls were classified as stunted and underweight. A similar number of girls were also classified as underweight and stunted compared to Indian cut-offs. So, the stunted and underweight estimates of WHO-specific relative child health measures may be misleading. Due to the changing pattern of population increase throughout time, it is recommended that growth references be updated regularly. The weight and height percentiles vary significantly across countries and city to city, particularly in childhood.

The study concludes that Indian growth standards significantly differ from WHO growth standards for height-based metrics, which have a significant intergenerational component. The average height is higher for Indian children compared with the WHO growth standards, although these differences are disappearing at younger ages. It is also found that healthy Indian infants are heavier than the WHO norms. So, this study suggests the WHO-specified relative child health measures stunted and underweight estimates may be reviewed and revisited. Consequently, building well-known countrywide curves is highly prioritized in health dietary and medical intervention guidelines. While establishing policy and program interventions to combat undernutrition, it is critical to consider the population's specific context, particularly

Ethics Approval and Consent to Participate

Ethics approval and participant consent were not necessary as this study is based on a secondary previously-published de-identified database NFHS. The NFHS survey received ethical approval from the International Institute for Population Science's Ethical Review Board and the Indian government. Participation in the survey was totally voluntary. Prior informed written consent was obtained from each respondent.

Consent for publication

N.A.

Availability of data and materials

The NFHS dataset used during the current study is available in the public domain from the Demographic and Health Surveys Program (DHS) website. https://dhsprogram.com/data/available-datasets.cfm

Competing interests

There are no competing interests.

Funding

No funding was received to conduct this study.

Authors’ contributions

KK conceptualized the paper and finalized the analysis plan; DP analyzed the Dataset, generated the tables, and prepared the first draft of the paper. Both authors read and approved the final manuscript.

Acknowledgment

N.A

Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, De Onis M, Ezzati M, Grantham-McGregor S, Katz J, Martorell R, Uauy R. Maternal and child undernutrition and overweight in low-income and middle-income countries. The lancet. 2013 Aug 3;382(9890):427-51.
International Institute for Population Sciences and ICF. National family health survey (NFHS-5), 2019-21: India. 2019 International Institute for Population Sciences. Available at http://rchiips.org/nfhs/NFHS-5_FCTS/India.pdf
Government of India. POSHAN Abhiyaan. Ministry of Women and Child Development. Government of India. Government of India, New Delhi. 2017. Available at http://poshanabhiyaan.gov.in/#/
World Health Organization. WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development. World Health Organization; 2006.
Panagariya A. Does India really suffer from worse child malnutrition than sub-Saharan Africa? Economic and Political Weekly. 2013 May 4:98-111.
Little MA. Evolutionary strategies for body size. Frontiers in Endocrinology. 2020 Mar 10;11:107.
Natale V, Rajagopalan A. Worldwide variation in human growth and the World Health Organization growth standards: a systematic review. BMJ open. 2014 Jan 1;4(1):e003735.
Cole TJ. Fitting smoothed centile curves to reference data. Journal of the Royal Statistical Society: Series A (Statistics in Society). 1988 May;151(3):385-406.
Cole TJ. The LMS method for constructing normalized growth standards. European journal of clinical nutrition. 1990 Jan 1;44(1):45-60.
Cole TJ, Green PJ. Smoothing reference centile curves: the LMS method and penalized likelihood. Statistics in medicine. 1992;11(10):1305-19.
Khadilkar VV, Khadilkar AV, Chiplonkar SA. Growth performance of affluent Indian preschool children: A comparison with the new WHO growth standard. Indian pediatrics. 2010 Oct;47:869-72.
Agarwal KN, Agarwal DK, Benakappa DG, Gupta SM, Khanduja PC. Growth performance of affluent Indian children (under-fives).1991.

Table 1: The power of a Box-Cox transformation (L), the median (M), and the coefficient of variation (S) for height

Age (in months)	Boys				Girls
Age (in months)	N	L	S	M	N	L	S	M
0	15	-3.16	0.14	56.63	19	1.75	0.20	53.92
1	41	-1.05	0.10	54.55	47	0.63	0.10	53.10
2	57	1.21	0.15	59.11	42	2.47	0.10	58.26
3	70	-1.36	0.11	61.47	53	2.52	0.08	59.35
4	78	-1.77	0.10	64.21	71	1.67	0.09	62.97
6	71	0.53	0.12	67.05	61	-0.32	0.09	66.36
9	70	0.96	0.07	71.65	56	-1.70	0.08	70.57
12	76	0.47	0.06	74.30	67	-0.96	0.07	72.54
15	72	-1.61	0.06	76.36	76	-2.90	0.07	75.37
18	88	-1.07	0.08	79.19	79	2.27	0.07	78.52
21	87	-2.30	0.07	80.61	56	2.89	0.06	81.13
24	65	-1.95	0.07	84.59	52	0.56	0.08	83.47
30	89	-2.64	0.07	89.58	78	-0.01	0.07	87.08
36	78	3.87	0.05	92.08	71	3.52	0.06	92.08
42	106	0.63	0.06	95.90	84	-0.85	0.07	94.39
48	76	1.87	0.08	101.03	64	-3.50	0.06	98.64
54	88	-2.33	0.07	103.09	87	-0.14	0.06	101.68
59	85	2.56	0.07	104.94	63	5.43	0.05	105.17

Table 2: The power of a Box-Cox transformation (L), the median (M), and the coefficient of variation (S) for weight

Age (in months)	Boys				Girls
Age (in months)	N	L	S	M	N	L	S	M
0	15	-0.84	0.30	3.48	19	1.07	0.18	3.20
1	41	0.88	0.19	4.20	47	-0.52	0.19	3.83
2	57	-0.27	0.29	4.95	42	0.94	0.19	4.69
3	70	0.62	0.23	5.76	53	-0.76	0.19	5.13
4	78	-0.40	0.24	6.24	71	-0.07	0.17	5.86
6	71	0.79	0.20	7.44	61	-0.14	0.19	6.64
9	70	0.46	0.14	8.23	56	-0.35	0.18	7.64
12	76	0.95	0.14	8.97	67	1.76	0.13	8.51
15	72	1.48	0.17	9.57	76	0.39	0.17	8.87
18	88	0.41	0.13	10.00	79	-0.96	0.14	9.42
21	87	-0.24	0.16	10.32	56	1.28	0.19	10.74
24	65	-0.76	0.13	10.65	52	1.12	0.19	10.61
30	89	-1.21	0.15	11.73	78	-0.83	0.17	11.36
36	78	-0.32	0.14	12.21	71	1.04	0.15	12.10
42	106	-0.24	0.14	13.34	84	-1.37	0.14	13.23
48	76	0.98	0.19	14.44	64	0.91	0.17	14.22
54	88	1.22	0.17	15.22	87	0.11	0.14	14.82
59	85	0.88	0.21	15.38	63	-0.87	0.15	14.93

Table 3: Descriptive values for calculated Z scores and WHO Z scores for individual children in India

Statistics	Calculated Z score	WHO Z score
Children Total	2,80,889 8,824
Selected for Reference Population	2,80,889 8,824
Height for Age
Mean	-0.48	-1.52
SD	1.11	1.68
Weight for Age
Mean	-0.48	-1.54
SD	1.05	1.22

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Does WHO child growth standards suitable for Indian Children: A comparison of anthropometric measurements from a cross-sectional survey

Status:

Version 1

Abstract

Figures

Background

Methods

Results And Discussion

Conclusion

Declarations

References

Tables

Additional Declarations

Status:

Version 1