Association of Novel Complementary Feeding Indicators with Length‐for‐Age Z‐Scores in Rural Malawi

Abstract

BackgroundInadequate complementary feeding (CF) from 6–24 months contributes to growth failure in many low income settings. However, recommended indicators of CF are inconsistently associated with length‐for‐age z‐scores (LAZ).ObjectiveTo assess the relationship between novel indicators of CF, including consistency and meal volume, with LAZ in rural Malawi.Methods1,331 households with a child aged 6 to 17 months were sampled from 53 communities in Mchinji district, Malawi using a single stage cluster sampling approach based on caregivers' participation in community women's groups. Data were collected on household sociodemographics, CF practices, and child anthropometry using WHO recommended methods. To assess food consistency, caregivers selected from a series of photographs that which most closely resembled their child's food. Caregivers estimated meal volume using uncooked rice which was then measured; meal volume was then classified as either below or above the age‐specific median for the following age categories: 6–8 months, 9–11 months, and 12–17 months. A summary feeding index (range 0–8 points) was created based on partially or fully meeting WHO recommended CF practices in addition to meal volume and consistency. Length‐for‐age z‐scores, derived from WHO Child Growth Standards, were regressed onto individual CF indicators and the summary feeding index, controlling for covariates. Logistic regression was used to assess the relationship between stunting and scoring in the second or third range of index scores (3–5 and 6–8 points, respectively).ResultsMean child age was 10.8 ± 3.2 months and 51% were female. There was no association between LAZ and minimum dietary diversity (P=0.4), minimum meal frequency (P=0.6), or minimum acceptable diet (P =0.8). There was evidence of interaction between child's age and both food consistency and meal volume. Among 6–8 month old infants, thicker food consistency was positively associated with LAZ (β = 0.14, 95% CI 0.02, 0.27); there was no relationship in older age categories. Meal volume was positively associated with LAZ in children 9–11 months (β = 0.29, 95% CI = 0.04, 0.54) and 12–17 months (β = 0.33, 95% CI = 0.10, 0.56), but not those aged 6–8 months. Each one point increase in the summary feeding index (median=5, IQR=3), was positively associated with LAZ (β = 0.05, 95% CI 0.00, 0.10) and reduced odds of stunting (OR = 0.93, 95% CI 0.87, 1.00). Compared to scoring in the lowest index category, scoring in the middle (3–5 points; OR=0.69, 95% CI 0.46, 1.05) or highest range (6–8 points; OR=0.58, 95% CI 0.38, 0.88) was associated with reduced odds of stunting.ConclusionNovel indicators of CF, such as consistency and meal volume, may be useful to understand the totality of CF and its relation to child growth in low income settings.Support or Funding InformationFunding and support for this research was provided by Concern Worldwide; Malawi Department of Nutrition, HIV and AIDS; World Bank; Emory Global Health Institute; Laney Graduate School, Emory University; and the National Institute of Diabetes and Digestive and Kidney Disease of the National Institutes of Health under award number T32DK007734.