August 29, 2022 | Erin Bluvas, firstname.lastname@example.org
Arnold School researchers Edward Frongillo and Alexander McLain have partnered with scientists from UNICEF and the World Health Organization, including WHO collaborator Elaine Borghi, to develop a method for estimating childhood stunting and overweight trends in the European region. The team’s research offers a solution to a long-time problem facing scientific and humanitarian efforts to understand and improve childhood growth and development. They published their findings in the Journal of Nutrition.
Frongillo is a professor of health promotion, education, and behavior and global nutrition researcher. McLain is an associate professor of biostatistics whose work specializes in solving problems that arise within biomedical studies, including methodological challenges such as missing data. Their combined expertise enabled them to estimate childhood growth trends in an area with a sparsity of data.
“Global rates of childhood stunting, which is a result of chronic malnutrition and inequality, are still unacceptably high, and childhood overweight and obesity is increasingly common,” Frongillo says. “It’s important to monitor countries’ progress toward achieving nutrition targets, such as a reduction in stunting, but lack of data and incomplete data makes monitoring trends challenging.”
“Data sparsity is a common challenge in nutritional epidemiology and public health,” McLain adds. “But new developments in statistical methods now allow us to gain valuable insights using longitudinal data. We implemented statistical methods to extract useful information on child malnutrition trends from sparse longitudinal data.”
For this project, the researchers compiled data related to childhood stunting and overweight prevalence from 26 countries within the WHO European region over a 30-year period. The databases were derived from sources such as national surveys and studies on childhood nutrition.
As expected, data was often limited and/or incomplete. For example, only 17 countries had stunting data from three or more surveys over the entire time period (only 18 countries for overweight). Further, few data sources existed for the years between 1990 and 1994. However, the researchers were able to use their modeling to predict prevalence estimates for countries and years where data was missing.
“Deriving useful information on the trends of childhood stunting or overweight from sparse longitudinal data is a useful exercise that can be repeated for other regions that aim to monitor trends in their countries’ levels of childhood malnutrition despite sparse data,” Borghi says. “Assessing these trends can provide important information to policy makers as they examine the effectiveness of nutrition programs over time or identify priority areas for action.”
The usefulness of this new method has already caught the attention of other researchers in the field. The Journal of Nutrition published an editorial on the study, highlighting its policy significance and future applications. The American Society for Nutrition reported on both the study and the editorial.