How to decode health hype, which often starts in the press release

Key takeaways

• In June 2015, researchers found 97 superlatives such as “miracle” and “revolutionary” used for 36 cancer drugs in the news; about half of those drugs lacked FDA approval, and 14% had never been tested in people.
• A study of 462 university press releases found that about a third exaggerated their own research, and news stories based on those releases were far more likely to exaggerate too.
• In an online experiment, adding the word “breakthrough” to a factual drug description doubled the share of readers who rated the drug very effective.
• Under federal rules, only a drug can claim to treat or cure a disease, and the FTC says soft words like “may” and “promising” don’t excuse weak evidence.
• Before you act on a hyped claim, ask who was studied, how, what changed and who is doing the talking.

In June 2015, two researchers spent five days searching Google News. They paired the phrase “cancer drug” with 10 superlatives, including “breakthrough,” “miracle,” “cure,” “marvel” and “home run”¹.

Their haul: 94 stories, from 66 outlets, with 97 superlatives aimed at 36 drugs¹. About half of those drugs had no approval from the Food and Drug Administration (FDA). Fourteen percent had never been tested in a human being¹.

The study, by Matthew Abola of Case Western Reserve University School of Medicine and Vinay Prasad of Oregon Health & Science University, was published online in JAMA Oncology in October 2015¹². Journalists supplied 55% of the superlatives and physicians 27%³. Industry experts, patients and one member of Congress chipped in too¹.

“We’ve been curing mice of cancer since the 1970s,” Prasad told NPR². A cure in a lab animal is a long way from a cure in you.

This library’s article on miracle-cure scams covers the red-flag language of outright fraud. This piece is about a quieter problem. The same vocabulary of hype turns up in respected places, from science journals to university press offices to the FDA’s own announcements. Once you know how it works, you can decode it anywhere.

Hype words show up in respected places, not just scam ads

The FDA has long warned shoppers about health-fraud phrases. A 2011 FDA consumer guide listed “scientific breakthrough,” “new discovery,” “quick fix” and “all natural, miracle cure” among its tip-offs to a rip-off⁴.

Those words aren’t limited to late-night ads. Researchers in the Netherlands led by Christiaan Vinkers counted 25 upbeat words, such as “novel,” “strong,” “unprecedented” and “promising,” in research abstracts from 1974 to 2014⁵⁶. They searched PubMed, the National Library of Medicine’s index of medical research.

The combined frequency of those words was 2.0% in 1974 to 1980. By 2014 it was 17.5%, a nearly ninefold rise⁵. Some words, including “novel” and “unprecedented,” rose by as much as 15,000%⁵. Neutral words didn’t show the same jump, and neither did upbeat words in books⁵.

The authors suggested that scientists may be marketing their results to stand out under heavy pressure to publish⁶. Their point was simple. When every finding is “novel,” the word stops telling you anything.

Exaggeration often starts before a reporter sees the story

It’s easy to blame journalists. A 2014 study in The BMJ suggests the problem often starts earlier.

A team led by Petroc Sumner at Cardiff University compared 462 health-related press releases from 20 leading UK universities, all issued in 2011, with the research papers behind them⁷. They also gathered 668 news stories based on those releases⁷.

Many press releases went beyond the research⁷:

• 40% gave more direct health advice than the paper did.
• 33% turned a correlation, two things that move together, into a claim that one causes the other.
• 36% of releases about animal research implied the results applied to people.

The news followed the release. When a release exaggerated a cause-and-effect claim, 81% of the stories based on it did too, compared with 18% when the release was accurate⁷. For animal studies, the figures were 86% versus 10%⁷.

This is the hopeful part. Exaggerated releases were not more likely to get news coverage⁷. In 2019, the Cardiff group tested this with a randomized trial, in which items are assigned by chance to one version or another. Nine UK press offices at journals, universities and research funders took part, with 312 press releases⁸.

Releases rewritten so their claims matched the evidence got news coverage just as often: 55% in both groups⁸. Adding plain caveats didn’t appear to hurt coverage either⁸. Hype, in other words, is a choice. It isn’t the price of attention.

One word can change what you believe a drug does

Does a single word matter? Researchers at Carnegie Mellon University and Dartmouth tested it. They showed about 600 adults online a short, factual description of a newly approved drug for advanced lung cancer⁹¹⁰.

Some people saw only the facts. Others saw the same facts plus the word “breakthrough” or “promising”⁹.

Only 11% of the facts-only group rated the drug “very” or “completely” effective. That rose to 23% with “breakthrough” and 25% with “promising”⁹. More people also judged the evidence strong: 59% and 63%, versus 43%⁹.

Worse, 31% of the “breakthrough” group wrongly believed the drug had been proven to save lives⁹. Asked to choose between two drugs for a potentially deadly illness, 92% picked the one called a breakthrough⁹.

This 2015 study in JAMA Internal Medicine matters because “breakthrough” is an official FDA label. Breakthrough Therapy designation goes to a drug for a serious condition when early clinical evidence suggests it may work much better than existing treatment on at least one important measure¹¹. The designation speeds up development and review. It is not an approval, and it is not proof¹¹.

That measure can be a stand-in, called a surrogate endpoint, rather than a direct sign that people live longer or better¹⁰¹¹. When the study was done, the FDA used “promising” in about half its press releases on breakthrough drugs¹⁰.

“Cure,” “treat” and “supports” are legal words with different jobs

In U.S. law, the verb on a label tells you what kind of product you’re holding.

Only drugs may legally claim to diagnose, treat, cure or prevent a disease¹². Dietary supplements can make “structure/function” claims about normal body function, such as “calcium builds strong bones”¹². The FDA does not approve those claims in advance. The company must have support for the claim, notify the FDA within 30 days of marketing and print a disclaimer saying the FDA hasn’t evaluated it¹².

The Federal Trade Commission (FTC), which polices advertising, adds two rules that matter for hype. First, health claims need “competent and reliable scientific evidence,” which the FTC says generally means randomized controlled trials in people¹³. Second, soft words don’t rescue weak science. The agency’s 2022 guidance says that saying a product “may” help, or calling results “promising” or “preliminary,” doesn’t warn shoppers that the evidence is thin¹³.

The FTC also says pictures can make claims that words avoid. In one of its examples, a “before” photo of an older woman with a walker and an “after” photo of her dancing likely imply that a joint ointment can dramatically ease arthritis¹³. This series’ companion article on before-and-after photos covers that trick in depth.

A decoder for the words of hype

Use this table on any headline, ad or press release. The left column is the word you see. The right column is the question that cuts through it.

Phrase you see	What it often means	What to ask
“Breakthrough”	An FDA designation based on early evidence, or just an adjective	Has it been approved for this use? Better than what?
“Revolutionary,” “transformative,” “miracle”	An opinion, often from a reporter or someone with a stake	Who said it, and do they gain if you believe it?
“Cure”	A strong medical claim that needs long follow-up; not allowed on a supplement	Cured in cells, animals or people? For how long?
“Kills cancer cells,” “reverses aging”	Often a lab-dish or animal result	Was it tested in people?
“Linked to” turning into “causes” or “prevents”	A pattern in data dressed up as cause	Was there a randomized trial?
“Promising,” “may help,” “early results”	Small or preliminary studies	How many people, and was there a comparison group?
“Clinically proven”	Should mean the product itself was tested	Which study, with this product and this dose?
“Novel,” “unprecedented,” “strong”	Common and rising science-speak	What do the actual numbers show?
“Supports,” “boosts,” “promotes”	Label wording the FDA doesn’t approve in advance	What evidence does the seller hold?
“Secret ingredient,” “new discovery,” “ancient remedy”	Classic health-fraud tip-offs	See this library’s miracle-cure scam checklist

How hype hurts: false hope, lost time and tuning out

The lung cancer experiment shows the first harm: false confidence. Among people told a drug was a breakthrough, 31% believed it was proven to save lives⁹. Among people who instead read the FDA’s required labeling language, which said continued approval depended on further trials, 10% to 16% did⁹¹⁰.

The second harm is lost time and money. A “cure” story can pull someone with a serious illness away from treatment that works. This library’s article on miracle-cure scams documents what that can cost.

The third harm is slower and harder to see. Rebekah Nagler of the University of Minnesota studied 631 U.S. adults surveyed in 2010 about news on wine, fish, coffee and supplements¹⁴. More than 71% said they heard moderate or high levels of conflicting information¹⁴.

People who heard more of it were more confused¹⁴. Confusion went along with backlash, a feeling that nutrition advice in general can’t be trusted. Backlash, in turn, was tied to slightly weaker plans to exercise and eat fruits and vegetables¹⁴.

That was a survey, so it shows a link, not a cause. A 2023 University of Pennsylvania dissertation found in an experiment that conflicting messages raised confusion and backlash, but effects on people’s plans were mixed¹⁵. Still, the pattern is worth watching. Today’s hype can set up tomorrow’s whiplash.

Your values already make you a good hype detector

Many natural-health readers are wary of drug-company marketing. That instinct is useful. The research above says to apply it evenly.

The same words that oversell a cancer drug also oversell a supplement, a detox kit or a wellness protocol. “Supports immune health” is legal label wording, not evidence¹². “Ancient remedy” may describe a real tradition, but the FTC says traditional use shouldn’t be advertised as a treatment or cure for a serious disease¹³.

Honest traditions describe how a plant has been used. Honest science describes what a study found and how sure we are. Both sound calmer than hype. Respect for the body’s own healing helps too, because some “miracles” are recoveries that would have happened anyway.

Some breakthroughs are real. NPR’s report on the superlatives study pointed to imatinib (Gleevec), trastuzumab and rituximab as cancer drugs that earned the word². The goal is to make bold claims show their evidence.

What the research shows about hype

Claim	Evidence level	Why
Exaggerated health news often begins in press releases	Strong evidence	A comparison of 462 releases and 668 stories, plus a randomized trial showing news follows releases78
Scientific writing has grown more upbeat	Strong evidence	A count of PubMed abstracts over four decades; the reasons are less certain5
Careful, accurate releases lose news coverage	Evidence says no	Neither the 2014 study nor the 2019 trial found a coverage penalty78
“Breakthrough” and “promising” inflate how well people think a drug works	Promising but early	One randomized online experiment with about 600 adults9
Conflicting headlines make people tune out all health advice	Promising but early	A survey link; an experiment showed mixed effects on plans1415

When to get help, and where to report

• Before you change a treatment. If a headline makes you want to start, stop or swap a treatment, talk with your clinician or pharmacist first. Bring the article with you.
• A product that claims to cure a disease. Report the ad to the FTC by phone at 1-877-FTC-HELP (1-877-382-4357)⁴ or online at ReportFraud.ftc.gov. If the website is down, your state attorney general’s office also takes consumer complaints.
• A product that harmed you. Tell your clinician and report the problem to the FDA⁴. In an emergency, call 911.

Questions readers ask

The bottom line

Hype words aren’t proof of fraud, and they aren’t proof of anything else. They show up in scam ads, research abstracts, university press releases and government announcements. Research shows they change what readers believe, and that careful wording costs scientists little or nothing in attention.

When you meet a “breakthrough,” a “miracle” or a “cure,” slow down and translate. Ask who was studied, how, what changed and who is speaking. The answers, not the adjectives, tell you whether a claim deserves your hope and your money.

Related reading

• How miracle-cure scams borrow the language you trust
• Why supplement labels say ‘supports’ and never say ‘cures’
• Chocolate tracks with Nobel prizes, and other lessons in telling a link from a cause
• Only 1 in 20 treatments tested in animals wins approval for people
• A before-and-after photo can be real and still mislead you

Sources

1. Reuters Health. “Glowing Terms Often Used for New Cancer Drugs in Health News” (on Abola MV, Prasad V, “The Use of Superlatives in Cancer Research,” JAMA Oncology, published online October 29, 2015, doi:10.1001/jamaoncol.2015.3931). Republished by Fox News, October 30, 2015. foxnews.com
2. NPR. “If A New Cancer Drug’s Hailed As A Breakthrough, Odds Are It’s Not.” KRCU Public Radio, October 29, 2015. krcu.org
3. The Lund Report. “Media’s Frequent Use of Superlative Terms to Describe New Cancer Drugs Questioned in Study.” October 30, 2015. thelundreport.org
4. U.S. Food and Drug Administration. “Health Fraud Scams…are Everywhere. Get the Facts.” HHS/FDA, November 2011. fda.gov
5. Vinkers CH, Tijdink JK, Otte WM. “Use of Positive and Negative Words in Scientific PubMed Abstracts Between 1974 and 2014: Retrospective Analysis.” BMJ, 2015;351:h6467 (record at Vrije Universiteit Amsterdam). research.vu.nl
6. Retraction Watch. “Do Science Findings Feel More Novel, Robust? They Are, at Least in Language.” December 15, 2015. retractionwatch.com
7. Sumner P, Vivian-Griffiths S, Boivin J, et al. “The Association Between Exaggeration in Health Related Science News and Academic Press Releases: Retrospective Observational Study.” BMJ, 2014. doi.org
8. Adams RC, Challenger A, Bratton L, et al. “Claims of Causality in Health News: A Randomised Trial.” BMC Medicine, 2019. link.springer.com
9. LabMedica. “Words That Inappropriately Enhance Perception of New Drug’s Effectiveness” (on Krishnamurti T, et al., JAMA Internal Medicine, published online September 21, 2015). October 7, 2015. labmedica.com
10. Reuters Health. “‘Breakthrough’ Drug May Not Mean What You Think.” Republished by Fox News, 2015. foxnews.com
11. U.S. Food and Drug Administration. “Breakthrough Therapy.” Content current as of January 4, 2018. fda.gov
12. U.S. Food and Drug Administration. “Structure/Function Claims.” Content current as of March 28, 2024. fda.gov
13. Federal Trade Commission. “Health Products Compliance Guidance.” December 2022. ftc.gov
14. Health Behavior News Service. “Contradictory Nutrition News Creates Consumer Confusion” (on Nagler RH, Journal of Health Communication, 2014). Newswise, January 28, 2014. newswise.com
15. Jesch EA. “Cascade of Confusion: Examining Incongruent Spillover Effects From Exposure to Conflicting Health Information Across Two Methods.” PhD dissertation, University of Pennsylvania, 2023. repository.upenn.edu