Back to browse results
Heterogeneity in adult anthropometry by socioeconomic factors: Indian National Family Health Survey 2006 and 2016
Authors: Rockli Kim, Praveen Kumar Pathak, Niharika Tripathi, and S. V. Subramanian
Source: European Journal of Clinical Nutrition, Published online; DOI: 10.1038/s41430-019-0511-0
Topic(s): Body Mass Index (BMI)
Women's health
Women's height
Country: Asia
Published: OCT 2019
Abstract: Background/Objectives Recent evidence suggests nonconstant nature of dispersion in adult women’s body mass index (BMI) across sociodemographic groups. The overall variances in BMI and height are also shown to have substantially changed over time. We modeled complex variation in adults’ anthropometry—BMI and height—by wealth and education, and assessed their differences over time in India. Subjects/Methods Data from a total of 768,130 women and 180,691 men from the Indian National Family Health Survey (NFHS) 2006 and 2016 were used for the analysis. The average association between wealth and education with anthropometry was assessed from linear regression models assuming constant variance. Individual heterogeneity was modeled to obtain separate variances in anthropometry for each wealth quintile and education level. All analyses were stratified by survey year and sex. Results On average, the positive socioeconomic gradient in adult’s BMI and height persisted over time with slight attenuation. The residual variance in BMI ranged from 10.1 to 14.9 (7.2–11.3) by education level and 6.1 to 17.4 (5.0–13.0) by household wealth for women (men) in 2006, and they increased over time for the lower socioeconomic groups but remained the same or decreased for the higher socioeconomic groups. No significant pattern was observed for variation in height for both genders. Conclusions We found potential reversal in the socioeconomic patterning in BMI variability in India as suggested by the increasing dispersion among the least educated and poorest populations. For a comprehensive understanding of nutrition transition in developing countries, it is necessary to assess the changes in means and variances of anthropometry in tandem.