Want to Fix Science’s Replication Crisis? Then Replicate

Want to Fix Science’s Replication Crisis? Then Replicate

Like
507
0
Wednesday, 19 April 2017
Opinion
This post was originally published on this site

Every month or so, the media reports on the latest so-called superfood that will either cure or cause cancer. As a scientist and lover of social media, I can recognize a whiplash phenomenon when I see one. The spotlight often shines brightest on the latest single research studies because the public finds new revelations most compelling, particularly if they challenge decades of established scientific research.

Exhibit A: a Cell Metabolism study from Qiao-Ping Wang and colleagues, published in the summer of 2016. The study observed that fruit flies ate more calories when sucralose, a low-calorie sweetener, was removed from their diet. Sounds plausible, right? After all, sucralose is intended to reduce the number of calories we consume. So it makes sense that fruit flies would eat fewer calories when sucralose is part of their diet.

The headlines from the study, however, went in a different direction: Some media outlets claimed sucralose can “cause” overeating. However, this is not what the totality of scientific evidence has to say on the subject. My colleagues and I at the IFIC Foundation gave our perspective of these findings and the reporting of the study. But now there’s more to the story (insert dramatic music and pause here), thanks to new findings, published in Cell Metabolism last month, from Jin Hong Park and colleagues.

The authors of the new study went old-school, rocking one of the primary principles of scientific inquiry: replication. In science, replication refers to the ability of an experiment or study to be repeated, either by the same group of scientists or others. It’s a novel concept these days, I know.

Using similar methodologies (you know, replication), the second study confirmed the findings of the first: Flies ate more calories while on non-sucralose diets than when fed diets containing sucralose. However, Park and colleagues found that the sucralose-fed flies were underfed, and that the subsequent “over-eating” was more likely to be a result of calorie deprivation than the impact of sucralose on neural responses. Turns out, the Drosophila melanogaster fly has an amazing ability to adjust food intake (in this case, reducing the number of calories eaten) based on perceived sweetness—a mechanism it uses to maintain energy balance because sweetness is often associated with calorie density.

If you were to read just these two studies and then attempt to conclude how sucralose might impact your appetite, food intake, or body weight, it would be a fool’s errand. But that’s where the existing body of literature can help: The totality of evidence repeatedly shows that low-calorie sweeteners like sucralose do not increase appetite, food intake, or body weight.

While the public often turns a cold shoulder to discussing the establishment of good scientific practices and guidelines, it’s a hot topic among scientists. In general, research either examines existing knowledge via replication and follow-up studies, or it expands on current knowledge via exploratory studies. Too often, exploratory studies are designed and reported to the public as cutting-edge and novel, yet in the same breath positioned as fact. When this happens, fewer scientists (and journalists) consider whether the results are clinically relevant, how the results align with similar research, or whether the methods were even appropriate. In other words, context gets cut from the conversation.

This doesn’t mean that cutting-edge, exploratory science is not important. Science advances thanks both to exploration and to replication and follow-up. However, since replication studies receive less funding and fanfare from the media and scientific communities, there seems to be little incentive to conduct them. I applaud Park and colleagues on their reproducibility effort, along with others who advance knowledge by conducting follow-up research. If we don’t confirm results, the question will always linger: How do we know it’s true?

Go Back to Top. Skip To: Start of Article.

Leave a Reply

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

Welcome to dev.jamair.tv