Changes in Dietary Variety Scores over Six Months Among 4- to 8-year-old Zambian Children

Abstract

Dietary variety scores are often used as indicators of dietary quality, especially among women and children in resource-poor settings. However, few studies have examined the stability of these scores over repeat measures. To study such change, we used data on dietary intakes of 4- to 8-year old children living in rural Zambia (n=188) collected by 24 hour recall interviews conducted six months apart. We calculated two dietary variety scores from each 24 hour recall: a 14-food group score and a count of individual foods. At both baseline in the late post-harvest season and follow-up in the early harvest season, the median food group score was 6 out of 14 (IQR (baseline and followup): 5,7) and the median food count was 8 (IQR(baseline): 6, 9.5; IQR(followup): 7, 10) . Mean absolute change from baseline to follow-up was 1.2 for the food group score (95% CI: 1.1, 1.4) and 2.3 for the foods count (95% CI: 2.1, 2.6). Agreement between tertile classifications at baseline and follow-up was poor (both scores: kappa