Skip to main content

Overblown coffee science undermined accuracy

By Audrey Goldfarb

When scientists make a discovery, journalists are tasked with communicating it to the general public. In biomedical research, bungling the message spreads health misinformation that can persist for decades. For example, the falsity that lengthening your DNA will make you live longer was perpetuated throughout the 2010s. A multi-million dollar industry was built on weak data marketed convincingly, and even strong evidence to the contrary failed to fully dismantle it. DNA lengthening products are still available in 2022.

To guide healthcare and lifestyle, biomedical science must satisfy a sufficient burden of proof. That is, hypotheses must be tested and supported by multiple experiments and in different contexts. When a single experiment is stated as truth, the public is led to draw premature conclusions.

A recent example of this occurred in June 2020 when scientists at the University of Bath warned against drinking coffee before breakfast. The researchers speculated about the health implications of their preliminary results, which the media translated as public health advice.

Coffee science

Worldwide, over 90% of adults consume some form of caffeine every day. Most Americans prefer coffee, and 90% of coffee drinking adults have at least one cup at breakfast.

Among caffeinated beverages, coffee seems like a good choice. Epidemiological surveys of thousands of people consistently show correlations between coffee consumption and better health. However, James Betts, Professor of Metabolic Physiology at the University of Bath, believes that this correlation is misleading. Coffee drinkers are more likely to be socioeconomically well-off and lead healthy lifestyles in general. “People erroneously assume there’s some protective effect of coffee,” Betts said.

Betts decided to investigate the metabolic impact of drinking coffee the morning after a bad night’s sleep. In 2020, he published his findings showing that following disrupted sleep, drinking coffee back-to-back with a sugary beverage led to a bigger spike in blood sugar.

A preliminary study

Betts’ study was designed to test the effects of poor sleep on the body’s ability to tolerate sugar. After one night of disrupted sleep, the researchers gave participants either coffee or hot water, followed by breakfast. They found that the coffee drinkers’ blood sugar was higher overall, indicating an impinged metabolic response.

High blood sugar after a meal could indicate a problem with carbohydrate tolerance. However, in this experiment, participants consumed neither a normal cup of coffee nor a typical breakfast. The dose of caffeine the researchers used was 300 milligrams, the equivalent of five shots of espresso and over double the average daily intake in the United States.

“Breakfast” was 65 grams of a sugar substitute that has a glycemic index double that of table sugar, meaning it makes blood sugar spike higher and faster. In one gulp, the study participants consumed nearly three times the maximum daily amount of sugar recommended by the American Heart Association. Proof-of-principle studies like these are important as a first pass, but to establish real-world conclusions, they must be bolstered by further investigation.

The blitz

 After the University of Bath issued a press release about the work, Betts and his co-author, Harry Smith, were met with a media storm. They were flooded with interview requests, and dozens of articles were published either overblowing or completely misstating their findings.

Drink coffee after breakfast, not before, for better metabolic control,” stated the University of Bath press release.

“A study has found that drinking coffee first thing can have a negative effect on blood sugar control – a risk factor for diabetes and heart disease,” CNN reported.

Another headline read, Scientists Find It Is Better to Drink Coffee After Breakfast, Not Before – Here’s Why.” Remember, Betts and his team only tested what happened when coffee was given before breakfast; they never measured the blood sugar of people who drank coffee after.

Betts said that in his experience with the media, accuracy was not a priority. He was misquoted in multiple articles and many of the headlines were misleading. “It was a bad experience,” Betts said. “It was sad that we spent a lot of time and money on learning these things and then it gets completely contorted by the media.”

Smith says the recommendation that people should avoid pre-breakfast coffee was overblown. “That’s not what we said, but it’s unavoidable, really,” Smith said. “It ends up being just [a game of telephone].”

The consequences

Unless I start swapping my morning oatmeal for Yoo-hoo, I won’t be skipping coffee. Smith still starts his day with espresso. If someone publishes a follow-up study that tests a normal breakfast with proper experimental controls, I’m sure we’ll both reconsider.

But while Smith and I are trained to interpret raw data, non-experts rely on journalists to explain the science in plain language. When journalists overblow data, they misrepresent  preliminary findings as definitive conclusions.

Journalists’ responsibility 

“There is a temptation to think that leaders in science have the answers to everything including health, and I think that makes things confusing,” Dr. Mary Armanios, M.D., Professor of Oncology at Johns Hopkins University School of Medicine, said.

In the lab, experiments are frequently artifactual and unintentionally misleading. A scientific paper is like one dot on a graph; you can’t tell if it’s an outlier until more dots fill in to form a pattern. How can the reader distinguish patterns from single dots?

“Everybody should be taught more critical thinking skills in school,” Betts said. “That would be the biggest jump for society, if people could analyze information for themselves.”

Frankly, it’s unrealistic to expect everyone to interpret data like a biomedical scientist. Scientific papers are notoriously dense with jargon. Even with rigorous training, interpreting data is hard work. Further, despite a movement towards open access publishing, much of this literature is still guarded behind paywalls that charge, on average, $32 for a single paper.

Journalists exist to bridge this gap and report scientists’ findings both accurately and in context, which also means being discerning. Not all studies with flashy claims are ready for the spotlight.

Your browser is out of date!

Update your browser to view this website correctly. Update my browser now