Growth Estimation of Under-Five Children Using Statistical Models in Central Region of India

Abstract

Background and aimsTo determine the suitability of 11 basic statistical models for estimating child-growth of under-five children and to bring-forth estimated growth curves for mean height & mean weight by their selected birth-weight categories for Central Region of India. MethodsThe study used fourth round of National Family Health Survey-4 (NFHS-4) data of India, consisting of 75,645 under-five children, belonging to 3 Indian States – Chhattisgarh, Madhya Pradesh & Uttar Pradesh. The children of the Region were first divided into 4 sub categories according to their birth-weight: (i) < 2000 gm, (ii) 2000–2499 gm, (iii) 2500–2999 gm (iv) 3000+gm, growth curve for mean height and mean weight were estimated for two sexes. ResultsThe significant association of 7 socio-demographic factors studied, namely – age & sex of child, birth-order, BMI, mother's highest level of education, place of residence and wealth index. Further, Cubic Model and Power Model, demonstrated best-fit to height & weight data of under-five children, belonging to different birth-weight categories, for estimating growth of boys & girls separately. These models enabled us to estimate mean height and mean weight, with 95% CI, for boys and girls separately by different birth-weight categories. ConclusionsStudy concluded that 7 socio-demographic factors were significantly associated with birth-weight. Further, Cubic Model and Power Model were most suitable for estimating child growth in terms of mean height & mean weight for boys and girls – considering specific birth-weight categories.