If you follow food news the way I do, then you’ve likely heard about Brian Wansink’s studies out of Cornell’s Food and Brand Lab. He’s been the director of the lab for years, and his studies have been published in premier medical journals like JAMA. He’s been regarded as an authority on nutrition, particularly in the area of how we interact with our food environment. He was given the title of “one of the world’s renowned eating behavior experts,” at least until recently.
One of his many studies involved people sitting at a table, spooning soup. What the subjects did not know was that as they were eating the soup, a system under the table was refilling the bowl. The idea was to see if despite the level of soup not going down, they would feel satiated and stop eating. Some subjects did, but many kept right on eating. The conclusion was that when we visually still see large amounts of food, despite being “full,” most of us continue to eat — we’ve essentially lost the ability to self-regulate food consumption. The theory makes sense but was this a validated study?
In September of 2018, six of his papers were retracted from reputable journals in one day, and in total, 15 have been called into question or retracted. Brian Wansink was found to have committed “academic misconduct” making up data or creating analysis results that simply were not proven by his research. You should know that 100 calorie snack packs essentially came to be, based on his research. The deepening investigation seems though, to suggest, that he engaged in serious data manipulation.
Wansink’s studies which have been cited more than 20,000 times, were called into question over the years by other career researchers. Prior to this research debacle he was appointed to lead the USDA committee on Dietary Guidelines in 2007-2009, he’d been key note speaker at countless food and nutrition conventions around the world, and he’s authored books that were well received. But it seems that his critics were correct — the findings he offered and his methods were seriously flawed. So, who can we trust when it comes to solid nutrition advice?
You should know that there is the temptation, unfortunately, to make studies sensational because of a “publish or perish” mentality. Wansink seemed to thrive on churning out loads of studies, with findings that would go viral. Some researchers may be tempted to find more significant findings than the studies actually reveal — a concept called p-hacking. Getting statistical results that are significant also means you are more likely to get published.
As a health writer, I certainly have “experts” whose work I follow. Most of these educated researchers have decades of published work that has withstood the test of time, in terms of peer vetting. Often, their research or commentaries present nutrition advice that is sensible, not sensational. Most of their presentations tend to be balanced, and they often question trendy studies that make it into the media. Their research and observations are often based on very large, long term data bases like NHANES. These experts don’t tend to make grandiose claims. They are also typically reluctant to abandon findings from large well-vetted studies, when smaller sensational studies that immediately percolate, seem to negate the research.
How do you know if a study’s findings are legitimate?
The researcher agrees to publicly pre-register the study design which helps to limit cherry-picking findings.
The researcher is willing to have his data scrutinized by other respected colleagues which means they are open to peer-review of their study.
- The researchers are open to replication of the study, meaning that they are willing to see if the study holds up to very close scrutiny.
These three features did not apply to Wansink’s approach to research.
If a diet seems too good to be true, it likely is. If it promises to help you lose significant amounts of excess weight, but can you sustain it? Will you re-gain all the weight and then some? Does it ask you to eat weird food combinations or abandon entire food groups? If you have a serious health condition then yes, you may have to go on a rigid diet for a period of time, but most nutrition experts will still have you eat a balanced variety with certain specific cautionary guidelines.
In terms of “expert” advice in the news, so many consumers typically buy into celebrities who have zero nutrition education. I get it. They look beautiful, so you believe following their recommendations will yield the same results for you. Mostly they won’t because the average person does not have the support team or financial advantage that these celebrities have. We also don’t have the time they spend adhering to diet preparation demands or the hours necessary for the type of exercise they commit to. Mostly it is not research-based advice.
I defer to people like Marion Nestle, an academician who has spent time at NYU and who is a highly regarded author (by her peers) in nutrition. I defer to Dr. David Katz, founding director of Yale University’s Griffin Prevention Research Center. They’ve both spent decades investigating food science and nutrition and their publications have withstood the test of time. They are also well-regarded by their peers. Dean Ornish has been a maverick focused on reversing heart disease, but his diet demands extraordinary commitment. Still, he has proven long term research results to back his methods.
Yes, Suzanne Somers looks great for her age, but she touts some rather alarming diet and hormonal advice. I take issue with her being branded as a lifestyle guru. On the other hand, I have no issue with a celebrity representing a diet program like Weight Watchers or Nutrisystem. These diet programs offer support, sensible food plans and allow individuals a broad range of diet options to fit their goals and lifestyles.
Beware the expert who offers bizarre, unsustainable food advice, and question the small studies that seem to negate established research findings. I’m always open to updated nutrition research, but it has to come from the right source, be based on a substantial pool of subjects, and have a well-vetted program design. You can always check information with your health professional. Believe it after they sign off on it!
Note: There are several articles on HeathCentral that mention Wansink studies. At the time these articles were written, his research had been well-regarded. There will likely be efforts to see if the studies can be replicated.