DEVELOPING MODELS TO PREDICT STUNTING AMONG 6-59 MONTHS CHILDREN IN A SLUM OF KOLKATA
Background: India is one among the many countries where child malnutrition is severe and is a major underlying cause of child mortality. 48 per cent of under five children in India are stunted and India accounts for more than 3 out of every 10 stunted children in the world. Even after implementation of the national nutritional programmes the magnitude of the problem remains at large, this necessitates a detailed analysis regarding the factors leading to stunting and to identify which group of factors to target first to make a significant effect.Methods: This Community based cross-sectional study was done in the service area of Urban Training Centre in Baghbazar, Kolkata with 84 children aged 6-59 months. Anthropometric measurements of the study population were done using standard methods and information regarding risk factors were collected from their parents using a pre-designed questionnaire. WHO Anthro for personal computers version3.2.2 (Z score) and SPSS version16.0 was used to analyse the data (multivariate logistic regression and ROC curve). Results: The study indicates a 26.2% prevalence of stunting among the study population. To observe the risk factors for stunting two models were compared by creating ROC curves, a socio demographic factor model which explained 70% of the stunting in the population in comparison to the second model regarding the factors related to birth and feeding practices which explained 65%. Conclusion: The current study further emphasizes the need to implement relevant interventions to combat malnutrition in this region and other similar settings.