Child anthropometry in cross-sectional surveys in developing countries: an assessment of the survivor bias

Abstract

In cross-sectional surveys, the sample of children with anthropometric measurements is not representative of all children in a birth cohort, since only children surviving to the survey date are measured. This survivor bias may have implications for studies of trends and differentials in anthropometric indicators. In this paper, the effects of the survivor bias on the estimates of child anthropometric indicators are assessed by 1) reviewing evidence from longitudinal studies on the prevalence of malnutrition among deceased children and among surviving children and by 2) analyzing retrospective data on child mortality and cross-sectional data on child anthropometry in 17 national surveys that are part of the Demographic and Health Surveys Program. It is concluded that comparisons of anthropometric data across geographic units, population subgroups, and calendar time are marginally affected by the survivor bias, unless mortality differences between the birth cohorts are very large (e.g., well over 50 per 1,000 births).

PIP: The effects of survivor bias in cross-sectional surveys of anthropometric measurements of children, that is overestimates due to omission of children that died before the survey date, were estimated by reviewing longitudinal studies for prevalence of malnutrition in deceased and surviving children, and analyzing retrospective data on child mortality and cross-sectional data on child anthropometry. Data were taken from 17 Demographic and Health Surveys, collected from 1986-1989. The odds ratio of the prevalence of malnutrition in deceased to that of living children was computed to assess how much worse the nutritional status of deceased children is to surviving children. In most countries 50% of the dead children were malnourished. The ratios were higher if malnutrition was more severe, specifically if the differences were pronounced for weight for age, and weight for height, or for severe stunting, below -3 SD from the median height for age. The proportion dead among children 3-36 months ranged from 2.5% in Sri Lanka to 15% in Mali, and was higher in older children. The largest bias was seen in weight for age: e.g., in Mali, where mortality is highest, the proportion of underweight increased by 4.6%, and of severely underweight by 2.7%. There was a relationship between level of malnutrition and child mortality at the country level, with considerable variation between countries, and a few outstanding outliers, countries whose malnutrition was much higher than expected from mortality rates. The effects of mother's education and of length of preceding birth interval were computed, and the odds ratio rose only from 1.8 to 2.2. It was concluded that comparisons of anthropometric data across countries and time are only marginally affected by survivor bias, unless the mortality differences between birth cohorts are very large, over 50/1000 births.