5 tips for writing better health news headlines

headlines-blog-editFellow journalists, we feel your pain: Writing headlines for health news is difficult.

Findings from medical research are often tricky to condense into a single, snappy and accurate sentence. Yet, it’s so important to get right, explains Sharon Dunwoody, a professor emeritus at the University of Wisconsin-Madison’s School of Journalism and Mass Communication, and a contributor of ours.

After all, the headline might be the only thing a person reads. And if people do click through, it’s the headline that leads readers to make inferences about what they will find in the story, she noted. “And those inferences then guide the subsequent interpretation of the text.”

With that in mind, we bring you 5 tips for writing better health news headlines–and we welcome hearing your experience and advice writing headlines, via the comments section.

1. Learn the difference between association and causality

To write better headlines, make sure you understand the narrative differences between association and causality. Here’s a quick primer–and why it matters:

Association: Many studies are observational or epidemiological. This means they examine associations between certain variables and outcomes, but they aren’t conclusive. For example, a study that asked people to recall their daily coffee intake and give a detailed medical history. Any patterns found there–such as non-coffee drinkers reporting more diabetes–show an association. For stories on these types of studies, your headlines should not imply cause and effect, that “coffee reduces diabetes risk,” for example.

Causality: When studies are randomized and controlled (such as including a placebo group) they are truly experimental and may be able to show cause and effect. For these studies, your headlines can be more conclusive.

2. Watch those “X may/might do Y” style headlines.

We’ve seen a lot of these recently. Science writer and HealthNewsReview.org reviewer Matt Shipman recently pointed out this one in a story review:

“Meditation may be the answer to relieving chronic back pain.”

“This might as well be replaced by ‘meditation may not be the answer,’” Matt wrote. “These sorts of headlines tell readers very little.”

Using “may” here suggests there’s some wiggle room in the findings, yet following it with “the answer” implies causality–but in a murky way.

In reality, these researchers were examining if meditation–together with yoga–were useful add-on treatments to standard medical care (including medication). Meditation was not being studied as a replacement, or, aka “the answer” for back pain. Nor was it the only intervention that the test group got, since it was meditation and yoga.

What might be a better alternative? First, get specific.
“Study participants who added meditation to treatment plan report more relief than control group”

And work from there, condensing it down while aiming for accuracy and clicky-ness.
“Meditation: Study shows it helped some back pain sufferers”

Another recently reviewed story included the headline “This New Blood Test Could Detect Concussions.”

According to the published study on which this story was based, this was preliminary research requiring more research to prove its usefulness, so we’d urge even more caution in the headline. Most likely the headline writer used “could” to indicate there’s still uncertainty; however, it’s too early to say that.

Better alternative? We liked this headline the AP used in its story about the research:
“Blood test for concussions? Researchers report some progress”

The headline introduces the topic and then suggests there’s an update to ongoing research on a blood test for detecting concussion. Readers who only read that headline have a very strong sense of where the research stands, and are unlikely to feel misled when they read the article.

3. Approach news release headlines with lots–and lots–of caution

In another recent example, this time from a news release, the headline read: “Eating Foods High in Vitamin C Cuts Risk of Cataract Progression by a Third.”

As typically happens, variations of this headline then appeared all over the internet after media outlets saw the news release and wrote about the study.

In our review we pointed out “the release includes an unsupported claim in the headline. This is an observational study — a type of research that can’t determine cause and effect. And so the active verb used here “cuts risk” is not appropriate to describe the results.”

Granted, this is a tough one to do quickly–and we welcome your brainstorming in the comments section. A better–though lengthier–alternative might be: “Study shows possible association between high Vitamin C diet and reduced progression of cataracts”

When is cause and effect OK? When writing headlines about a randomized trial–especially one involving thousands of patients and lasting several months or even years–you’ll have stronger evidence and data to point to in your headline.

4. Back up any grandiose claim in the headline (and the story) with evidence

Whether you’re writing about observational or randomized research the onus is on you to state the study’s evidence–or point out the lack of it. Anecdotes from experts or testimonials from patients are never enough. Take “first-of-its-kind” claims, notes Earle Holland, a science communications expert and contributor of ours.

Medical breakthrough

Always be cautious when using words like “cure,” “breakthrough,” “miracle,” and “first of its kind” in headlines.

“To my mind there is nothing wrong with labeling findings as ‘new’ but when the label is something like ‘the first of its kind,’ I want some indication in the copy that they can justify that,” Holland explains.

“For example, if a news release or news story headline was to say: ‘Researchers show for the first time that grass can be grown downward,’ it needs to be followed up with a statement that basically says, ‘Other research teams have previously tried to fool grass stalks into growing downward but have failed.’”

5. Try to avoid trap of “question mark journalism”

We see this one a lot, too. Writers try to skirt all these issues by simply adding a question mark to whatever bold statement they want to make–yet don’t want to be held accountable for. “Does your purse contain deadly microbes? Will coffee help you live longer?” And so on. Try to avoid these.


Resources from our toolkit:

Check out Observational Studies – Does The Language Fit The Evidence? – Association Versus Causation

You might also like

Comments (5)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Marc Beishon

April 7, 2016 at 12:17 pm

Reporters can suggest headlines but most are written by editors and subeditors who are not specialists so it’s no surprise that many get hyped up. It’s also the case that they may well be deliberately jazzed up into causality to justify a prominent position or as click bait – and the stories may be dropped otherwise. Sad but true.

    Gary Schwitzer

    April 7, 2016 at 12:34 pm


    Thanks for your note.

    We’re well aware of how and why headlines are created. No matter who is responsible, though, someone is, indeed, responsible. And that is part of our message.

    Gary Schwitzer

Ned Johnson

April 7, 2016 at 11:20 pm

While I support all of these cautions, the first one has been a pet peeve of mine since I took statistics as a college freshman. What we’re talking about here is called correlation in statistics, and absolutely does not imply causality. It simply asserts that two variables tend to vary in related ways. Yet not just headlines, but article text seems almost universally to use this false causality to prove points that they to be seen as true. I wrote a piece in 2009 on my blog entitled “The Parallax View of Thrill Seekers” in which I showed alternative interpretations of the findings reported in an article in New Scientist magazine about the relationship between a pair of proteins in the adrenal gland called ZERO and STREX. It was an attempt to show by example how equally reasonable interpretations can, but seldom are, made. For anyone who is interested, here is the address of that blog article: http://nedswords.com/the-parallax-view-of-thrill-seekers/.

Thanks for this article and for bringing these issues into the bright light of day. We can only hope that the right people read it, and respond appropriately.

Lawrence Kendall

April 8, 2016 at 10:07 am

Jon Stewart the originator of the Daily Show? Was Craig Kilborn the first host and originator? Should I even be asking the question?


    Joy Victory

    April 8, 2016 at 11:57 am

    Great catch Lawrence. You’re correct. I’ll edit that to simply say “host.”