Evaluation of Obesity Studies | Dr. Wendy Bennett
Articles Blog

Evaluation of Obesity Studies | Dr. Wendy Bennett

March 10, 2020

>>Hello, my name is Wendy Bennett and I’m an Associate Professor of Medicine at The Johns Hopkins
University School of Medicine and The School of Public Health. Obesity is an enormous
public health problem and it affects both children and adults. Many of our traditional interventions have focused on the
individual’s health behaviors, like diet and physical activity. So the motivation of this project was to understand what are the methods that researchers are using in these natural experiment
studies on obesity so that we can improve our understanding of the types of data sources we’re using, the data linkages, and
the different methods and analytic approaches being used. Of the 264 studies, we identified 53% of those were
natural experiment studies. And these again were studies where researchers did not have control over the intervention. Most of the studies that we
found were conducted in children and in the school setting. All of the natural experiment studies used a population based data source, and most of them used national level data or state level data. Fewer studies had access to
looking at community level data. Increasing when we find that studies that are assessing physical activity are using accelerometers or
other types of wearable devices to assess how active people are. In the natural experiment studies, the most common analytic approach was a cross-sectional comparison where the researchers compared those that were exposed to the policy or program or built environment change, versus those that were not
exposed to that change, or program or policy. So we found that natural
experiment studies had overall a high risk of bias and this was mostly due to
high losses to follow-up in their studies, as well as
high risks of confounding. For researchers, we
identified several areas where we think that researches can improve the types of data sources we’re using, the way we’re linking data, and increase the frequency of
linking data sources together. Policymakers are going to want
to use natural experiments and evaluations like them to understand whether
their policies and programs are effective in reducing obesity as well as preventing obesity. Obesity research, this
would be very helpful to understand quickly
which are the studies that are high quality and which are the most
effective strategies to reduce obesity and prevent obesity. All of this data that
we’re pulling together can be helpful in future studies when we’re trying to understand
what is most effective in our neighborhoods and
communities to impact obesity.

Leave a Reply

Your email address will not be published. Required fields are marked *