Related Experiment Videos
What proportion of children have a growth deficit?
1Department of Public Health and Epidemiology, Swiss Tropical Institute, Basel.
Insights
This study introduces a new statistical method to accurately estimate child growth retardation (lambda) from anthropometric data. The improved estimation helps in planning effective nutrition interventions and understanding population health variations.
Area of Science:
- Pediatric Nutrition
- Anthropometry
- Public Health Statistics
Background:
- Child malnutrition is often assessed using prevalence of standard deviation scores less than -2 (Pz < -2).
- However, Pz < -2 does not accurately represent the prevalence of growth retardation (lambda) and can obscure the distribution of growth deficits.
- Existing alternative estimators for lambda are known to be biased.
Purpose of the Study:
- To develop a statistically sound method for estimating the prevalence of child growth retardation (lambda) from cross-sectional anthropometric data.
- To provide nearly unbiased estimates of lambda and the distribution of growth deficits within a population.
- To offer a tool for better-informed public health planning and analysis of growth retardation.
Main Methods:
- A novel statistical procedure utilizing density ratios is described.
- This method estimates lambda using cross-sectional anthropometric data and international growth standards.
- The procedure also estimates the probability of individual child growth retardation.
Main Results:
- The proposed density ratio method provides nearly unbiased estimates of lambda.
- It accurately characterizes the distribution of growth deficits among growth-retarded children.
- The method was illustrated using data from the 1982/3 National Nutrition Survey of Papua New Guinea.
Conclusions:
- Accurate estimation of lambda is crucial for targeted public health interventions, distinguishing between severe malnutrition and widespread mild growth retardation.
- This method can inform decisions on whether to target interventions at the smallest children or the general population.
- Improved lambda estimates can help explain variations in the relationship between child morbidity/mortality risks and anthropometric indices across different populations.
Abstract:
Cross-sectional anthropometric survey data are frequently summarized to present estimates of the proportion of a child population who might be considered malnourished. A very commonly used summary statistic is the prevalence of standard deviation scores less than -2 (Pz < -2). Pz < -2 is not equivalent to the prevalence of growth retardation in the population, lambda. Identical values of Pz < -2 can arise as a result of a relatively low prevalence of predominantly severe malnutrition, or of universal or near-universal mild growth retardation. Alternative estimators which have been proposed give biased estimates of lambda. In this paper a simple statistical procedure using density ratios is described for estimating lambda from cross-sectional anthropometric data and international growth standards. This gives nearly unbiased estimates of lambda and of the distribution of the growth deficit in the growth retarded sub-set of the population. The analysis provides estimates of the probability that an individual child's growth is retarded. This is illustrated with data from the 1982/3 National Nutrition Survey of Papua New Guinea. Estimation of lambda for different indicators and for different populations could be useful in planning whether interventions should be targeted at the smallest children, or aimed at the population in general. The same estimates might also be used to help determine whether heterogeneity between countries in the reported associations between morbidity or mortality risks and anthropometric indices can be accounted for by differences in the amount of intrapopulation variation in growth retardation.