Revisiting The Stunting Metric For Monitoring And Evaluating Nutrition Policies

Contact: joanna.jia@wecistanche.com / WhatsApp: 008618081934791

Globally, as well as in India, the prevalence of stunting is a commonly used metric to quantify undernutrition among children. Stunting prevalence is the proportion of children younger than 5 years who are too short for their age, according to the 2006 WHO growth standard, constructed from the Multi Growth Reference Study (MGRS).1 Using the MGRS standard, around 38% of children in India are stunted.2 This metric is also used by countries to monitor the effectiveness of current policies that are in place to reduce the child undernutrition burden. In India’s flagship nutrition program, POSHAN Abhiyaan, launched in 2017, stunting is one of the monitoring and evaluation metrics, as are underweight, low birth weight, and anemia. As India formulates POSHAN Abhiyaan 2.0, the use of the stunting metric to measure the effectiveness of current interventions needs to be reconsidered.

desert ginseng cistanche for nutrition program

The growth of the children included in the MGRS reflects patterns that are to be expected under the ideal environment.3 A child is considered stunted if they are less than –2 SD from the median in a normalized distribution of height in the MGRS. By definition, the stunting prevalence among children living in ideal environments is expected to be around 2·3%. Therefore, having a standard provides a general sense of how far away countries such as India, with 38% stunting, are from the ideal.

Child height has a strong intergenerational component: children born to shorter mothers, on average, are shorter. Among the two dozen correlates that have been shown to be associated with the risk of stunting in childhood, maternal height is the strongest.4 The attained height of the mother, determined both by her own genetics and past nutrition-sensitive insults she might have endured, is then passed down to her children.5

As important as intergenerational exposures are for understanding child growth, current policies cannot do anything to alter the stature of the mother or the past deprivations experienced by the mother. Because the stunting metric automatically includes some degree of this intergenerational component, it is not appropriate as a measure to understand current interventions and progress. This point was well demonstrated in a multicountry study that compared crude stunting prevalence with prevalence standardized for maternal height. For India, the prevalence of maternal-height-standardized stunting was 25% compared with the MGRS-based crude prevalence of 38%.6 In a striking coincidence, a reduction to 25% prevalence was exactly the target reduction in stunting set by POSHAN Abhiyaan to be accomplished by 2022.

There is a major problem with using the MGRS standard for assessing children’s nutritional failure that has not been fully appreciated. In the case of India, the mothers in the MGRS, on average, were substantially taller than average Indian mothers. Consequently, inherent in the MGRS standard is not only an expectation of children growing up in an ideal environment but also the presumption of having taller mothers.

Ensuring ideal conditions while establishing a standard is understandable. However, including mothers who are substantially taller than typical Indian mothers in the standard complicates the use of stunting for assessing India’s current nutritional insults as the data will inevitably reflect the country’s shorter maternal stature. Therefore, stunting prevalence based on the MGRS standard will, to some degree, overestimate child undernutrition in shorter populations such as in India.7 In short, MGRS standards are unattainable, because the heights of mothers giving birth in the near future cannot be changed.

To put this in context, in the pooled-MGRS (consisting of samples from six countries, including India) the mean maternal height was 161·6 cm (and 157·6 cm in MGRSIndia).1 However, the mean height of mothers in India is 151·7 cm,6 which is 9·3 cm shorter than the pooledMGRS (and 5·9 cm shorter than MGRS-India). Notably, no single state or district in India matches the MGRSIndia’s maternal height, let alone the pooled-MGRS maternal height on which the child growth standards are based. This explains why, when maternal height is accounted for, India’s prevalence of stunting saw a downward revision of 13 percentage points, translating to about 15 million fewer children.6

The downward revision is observed across all states and union territories in India to varying degrees (appendix 1). In Uttar Pradesh, Bihar, Madhya Pradesh, and Rajasthan— all states with high stunting prevalence—there was a downward revision of 10 percentage points or more. Kerala and Goa had only marginal downward revisions. Furthermore, 603 of 640 districts had a downward revision of stunting prevalence after standardizing for maternal height (appendix 2). Areas with shorter mothers on the average and stronger correlation between maternal height and stunting will have greater downward revisions compared with areas where the mean maternal height was closer to the MGRS standard and the correlation between maternal height and stunting is weaker. The two methods of accounting for maternal height while estimating the prevalence of stunting (ie, using direct standardization [the standardized prevalence of stunting] and regression-based prediction [the predicted prevalence of stunting]) largely yield similar estimates (appendix 1, 2).6

These downward revisions do not necessarily mean that stunting and undernutrition are not problematic in India. Indeed, with a stunting prevalence of 25% (even after maternal-height standardization) it is clear that India is far from the expected ideal stunting prevalence of 2·3%. Indeed, it is likely that the underlying problem of child undernutrition has only worsened due to COVID-19 related disruptions.8

In summary, the stunting metric, due to its sensitivity to maternal height, cannot accurately assess the extent of the current undernutrition burden. Among the target indicators identified in POSHAN Abhiyaan, stunting is least indicative of current nutrition-specific and nutrition-sensitive conditions. Anthropometric-based measures of undernutrition, in general, are complex and lack the precision to accurately identify nutritional deprivation among children.9,10 As POSHAN Abhiyaan 2.0 is designed, a thoughtful deliberation on the choice of monitoring metrics that precisely and accurately capture India’s current nutritional challenges is urgently warranted.

*S V Subramanian, Omar Karlsson, Rockli Kim

Harvard Center for Population and Development Studies, Cambridge, MA 02138, USA (SVS); Department of Social and Behavioral Sciences (SVS) and Takemi Program in International Health (OK), Harvard University TH Chan School of Public Health, Boston, MA, USA; Division of Health Policy and Management, College of Health Science, Korea University, Seoul, Korea (RK); Interdisciplinary Program in Precision Public Health, Department of Public Health Sciences, Graduate School of Korea University, Seoul, Korea (RK)

Reference

1 WHO Multicentre Growth Reference Study Group. Enrolment and baseline characteristics in the WHO Multicentre Growth Reference Study. Acta Paediatr Suppl 2006; 450: 7–15.

2 International Institute for Population Sciences, ICF. India National Family Health Survey NFHS-4 2015–16. Mumbai: International Institute for Population Science and ICF, 2017.

3 Karra M, Subramanian SV, Fink G. Height in healthy children in low- and middle-income countries: an assessment. Am J Clin Nutr 2017; 105: 121–26.

4 Kim R, Rajpal S, Joe W, et al. Assessing associational strength of 23 correlates of child anthropometric failure: an econometric analysis of the 2015–2016 National Family Health Survey, India. Soc Sci Med 2019; 238: 112374.

5 Subramanian SV, Ackerson LK, Davey Smith G, John NA. Association of maternal height with child mortality, anthropometric failure, and anemia in India. JAMA 2009; 301: 1691–701.

6 Karlsson O, Kim R, Bogin B, Subramanian SV. Maternal height-standardized prevalence of stunting in 67 low- and- middle-income countries. J Epidemiol 2021; published online Feb 20.

7 Natale V, Rajagopalan A. Worldwide variation in human growth and the World Health Organization growth standards: a systematic review. BMJ Open 2014; 4: e003735.

8 Rajpal S, Joe W, Subramanian S. Living on the edge? Sensitivity of child undernutrition prevalence to bodyweight shocks in the context of the 2020 national lockdown strategy in India. J Glob Health Sci 2020; 2: e19.

9 Joe W, Rajpal S, Kim R, et al. Association between anthropometric-based and food-based nutritional failure among children in India, 2015. Matern Child Nutr 2019; 15: e12830.

10 Beckerman-Hsu JP, Chatterjee P, Kim R, Sharma S, Subramanian SV. A typology of dietary and anthropometric measures of nutritional need among children across districts and parliamentary constituencies in India, 2016. J Glob Health 2020; 10: 020424.