Effect of Differences in Month and Location of Measurement in Estimating Prevalence and Trend of Wasting and Stunting in India in 2005-2006 and 2015-2016

Emily M Madan; Edward A Frongillo; Sayeed Unisa; Laxmikant Dwivedi; Robert Johnston; Abner Daniel; Praween K Agrawal; Sila Deb; Ajay Khera; Purnima Menon; Phuong H Nguyen

doi:10.1093/cdn/nzaa092

Effect of Differences in Month and Location of Measurement in Estimating Prevalence and Trend of Wasting and Stunting in India in 2005-2006 and 2015-2016

Curr Dev Nutr. 2020 May 23;4(6):nzaa092. doi: 10.1093/cdn/nzaa092. eCollection 2020 Jun.

Authors

Affiliations

¹ Division of Nutritional Sciences, Cornell University, Ithaca, NY, USA.
² University of South Carolina, Columbia, SC, USA.
³ International Institute for Population Sciences, Mumbai, India.
⁴ UNICEF, New Delhi, India.
⁵ Ministry of Health & Family Welfare, New Delhi, India.
⁶ Poverty and Health Nutrition Division, International Food Policy Research Institute, Washington DC, USA.

Abstract

Background: Child undernutrition in India remains widespread. Data from the National Family Health Survey 3 and 4 (NFHS-3 and NFHS-4) suggest that wasting prevalence has increased while stunting prevalence has declined.

Objective: The objectives of this study were to do the following: 1) describe wasting and stunting by month of measurement in India in children <5 y of age in NFHS-3 and NFHS-4 surveys, and 2) test whether differences in the timing of anthropometric data collection and in states between survey years introduced bias in the comparison of estimates of wasting and stunting between NFHS-3 and NFHS-4.

Methods: Data on wasting and stunting for 42,608 and 232,744 children aged >5 y in the NFHS-3 and NFHS-4 survey rounds were analyzed. Differences in the prevalence of wasting and stunting by month of year and by state of residence were examined descriptively. Regression analyses were conducted to test the sensitivity of the estimate of differences in wasting and stunting prevalence across survey years to both state differences and seasonality.

Results: Examination of the patterns of wasting and stunting by month of measurement and by state across survey years reveal marked variability. When both state and month were adjusted, regardless of the method used to account for sample size, there was a small negative difference from 2005-2006 to 2015-2016 in the prevalence of wasting (-0.8 ± 0.6 percentage points; P = 0.2) and a negative difference in stunting prevalence (-8.3 ± 0.7 percentage points; P < 0.001), indicating a small bias for wasting but not for stunting in unadjusted analyses.

Conclusions: State and seasonal differences may have introduced bias to the estimated difference in prevalence of wasting between the survey years but did not do so for stunting. Future data collection should be designed to maximize consistency in coverage of both time and place.

Keywords: month of year; seasonality; stunting; undernutrition; wasting.