"Linked to" is not "causes": how to read health headlines
Health Insights · Research Explained
By Gabriella Langholz · October 2026 · 5-minute read
Key takeaways
- "Linked to" usually means researchers found a pattern, not proof that one thing causes another.
- Exaggeration often starts before the news story, in the press release.
- A large relative risk can mean a very small change in your actual risk.
- Studies in animals or cells are early steps, not advice for humans.
"Coffee linked to longer life." "Late nights linked to heart disease." Headlines like these appear every week. Some are based on solid research, others on early or limited findings. A few simple questions can help you tell the difference.
What does "linked to" really mean?
Many health headlines describe observational studies. In these studies, researchers observe large groups of people and look for patterns, for example whether people who drink more coffee have fewer heart problems.
Finding a pattern is called an association. It does not prove that one thing causes the other. People who drink coffee may also differ in income, sleep, smoking or overall health, and any of these could explain the result. Researchers call these hidden explanations confounding factors.
In 1965, the British epidemiologist Sir Austin Bradford Hill described a set of questions that scientists still use to judge whether an association is likely to be causal. They include how strong and consistent the link is, whether the exposure came before the outcome, and whether more exposure leads to more effect.1
Where does the exaggeration come from?
It is easy to blame journalists, but the picture is more complicated. A study published in the BMJ examined 462 press releases from 20 leading UK universities. It found that 33% of them made stronger causal claims than the research itself, and 40% gave exaggerated health advice. When a press release exaggerated, news stories were much more likely to repeat the exaggeration.2
A review of the health research stories most widely shared on social media in 2015 found that 48% of media articles, and 34% of the original academic papers, used language that was too strong for the evidence.3
Five questions to ask before you believe a headline
1. What kind of study was it?
Randomized controlled trials, in which people are randomly assigned to a treatment or a comparison group, are better at showing cause and effect than observational studies. Systematic reviews that combine many studies are stronger still.
2. Was it done in people?
Many promising results in cells or animals never work in humans. If the study was in mice, the headline should say so.
3. How many people took part?
A study of 30 people can be a useful start, but its results are less reliable than those from thousands of people.
4. Is it relative or absolute risk?
"Cuts your risk by 25%" sounds impressive. But if a disease affects 4 in 1,000 people, a 25% reduction brings it down to 3 in 1,000, a difference of just 1 person in 1,000. Researchers have shown that relative numbers like these are often misunderstood by patients, journalists and even doctors.4
5. Does it fit with other evidence?
One study rarely changes medical advice. Guidelines change when the evidence from many studies points in the same direction.
Words to watch for
- "Linked to," "associated with": a pattern, not proof of cause
- "May," "could": the result is uncertain
- "Breakthrough," "miracle," "cure": be extra careful
- "Doubles your risk": ask what the risk was to begin with
Does this mean observational studies are useless?
Not at all. Observational research is how we learned that smoking causes lung cancer, and it is often the only ethical or practical way to study long-term health effects. The point is not to dismiss these studies, but to read them for what they are: an important piece of a larger puzzle.
What should I do with a new health headline?
Before you change what you eat, stop a medication or buy a supplement, look for the original source, check whether trusted health organizations have said anything about it, and talk to your doctor or pharmacist. Good health decisions are built on the whole body of evidence, not on a single headline.
This article provides general health information and does not replace advice from your own doctor.
References
- Hill AB. The environment and disease: association or causation? Proc R Soc Med. 1965;58(5):295–300. https://doi.org/10.1177/003591576505800503
- 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;349:g7015. https://doi.org/10.1136/bmj.g7015
- Haber N, Smith ER, Moscoe E, et al. Causal language and strength of inference in academic and media articles shared in social media (CLAIMS): a systematic review. PLoS One. 2018;13(5):e0196346. https://doi.org/10.1371/journal.pone.0196346
- Gigerenzer G, Gaissmaier W, Kurz-Milcke E, Schwartz LM, Woloshin S. Helping doctors and patients make sense of health statistics. Psychol Sci Public Interest. 2007;8(2):53–96. https://doi.org/10.1111/j.1539-6053.2008.00033.x