No scientist ever says, “I’m making a non-evidence-based solution.” So, why is there so much discussion about evidence-based research in nutrition?
The evidence-based movement is about 20 years old, Dennis M. Bier, MD, director of the USA Children’s Nutrition Research Center and professor of pediatrics at Baylor College of Medicine in Houston, said Tuesday in the Rhoads Lecture. “It’s reached the stage of implementation pretty much everywhere.” Though, in the nutrition community, he said, people have been dragged “kicking and screaming.”
Are the standards of evidence in nutrition different than in other fields, he asked. Generally speaking, no. There remain expectations of:
- The explicit question
- The explicit outcome
- Intervention
- Replication
- The ability to replicate until it’s predictable
What’s potentially different, however, is whether or not you can actually carry out your study.
“But that doesn’t mean you suddenly elevate the level of evidence you have from C to A or from 6 to 2,” Bier said. “It means the final answer is maybe.”
Bier, who is also editor of the American Journal of Clinical Nutrition, pointed to the benefit of randomized control trials (RCTs) over observational studies. Observational studies, he explained are biased by definition.
To help illustrate the deficits of observational studies, Bier pointed to 14 studies in which the RCT didn’t support the results of the corresponding observational study. In fact, there were four cases in which the RCT even showed a negative, harmful outcome.
Because of issues with nutrition, such as long-term compliance, randomized controlled trials are challenging. Larger trials are better, however, as they tighten the precision.
Further complicating matters for nutrition researchers are several barriers to evidence-based work. For example, getting reliable data energy intakes is a significant, known challenge.
“Nutrition is the science of food, but we can’t reliably measure food,” Bier said. “This is an essential problem in the field of nutrition if we’re going to move forward.”
And perhaps one of the greatest challenges is perception in our reporting. We have, Bier said:
- Under-reporting of negative studies
- Over interpretation of negative/marginal results by authors
- Distortion of reported results by opinion leaders and the media
So, what can we do?
First, Bier advocates the sharing of data.
In addition, he suggests asking a number of questions (i.e., with 61 questions, you have a 95 percent chance of a “positive”).
He also advocates study registration, which has both statutory and transparency benefits. “No one should do a trial today that they don’t register because there will be journals who won’t take it,” he added, noting that many people actively avoid registering observational studies.
Bier recommends that all studies—including observational studies—should be registered, declaring endpoints and documenting any changes to the protocol or endpoints.
Maintaining standards of evidence-based practice in nutrition ensures a better reputation for the field and keeps us moving in the right direction, Bier said.
