Continued problems with inconsistent & adjacent news releases from BMJ

BMJ  publishes more than 50 specialist journals, and distributes news releases from these journals as well.

That’s a big workload, and sometimes it shows.

This week, BMJ sent out two releases in one email.

• From the British Journal of Sports Medicine, a news release headlined, “Half an hour of physical activity 6 days a week linked to 40% lower risk of death in elderly men.” 

The good:  The news release included this important caveat: “This is an observational study so no definitive conclusions can be drawn about cause and effect, and the researchers point out that only the healthiest participants in the first wave of the study took part in the second wave, which may have lowered overall absolute risk.”

The bad:  Despite that caveat, the news release used this subhead under the bold headline above:  “Impact on health as good as giving up smoking, suggest researchers.”

Wait a minute.  You can’t have it both ways. In one line you admit, appropriately, that cause and effect has not been established.  But up at the top, you allow researchers to get away with a huge, powerful cause-and-effect claim.

• From the Journal of Epidemiology & Community Health, a news release headlined, “Single motherhood before age of 50 linked to poorer health in later life.” 

The good:  “linked to” in the headline is OK, because, again, this was an observational study.  A link, or a statistical association, is all that was shown.

The bad:  Even though this was another observational study, as was the one above, and even though the news release appeared in the same email as the one above that included the caveats about the limitations of observational studies, this one did not.

We’ve been criticizing BMJ news releases for language used about observational studies for a long time.  Then we praised them when we first saw improvement. Then we noted a setback.  Now we’re stuck in limbo with mixed messages within one BMJ email.

At one point, a BMJ staffer wrote, “we sometimes wonder why you don’t seem to pick up issues with others’ press releases to the same extent.”

We’re not picking on BMJ.  Just yesterday we criticized a news release by the Lancet for a similar flaw.

Why does it matter?  Because I can guarantee you that news stories will follow the lead of these news releases and miscommunicate to the general public. We saw yesterday how The Lancet’s news release, “Testing hand-grip strength could be a simple, low-cost way to predict heart attack and stroke risk,” undoubtedly misled many journalists into reporting less-than-helpful stories.  And it’s again after these news releases from BMJ. See a search of news stories on the first release….and a search of news stories on the second release.  Good luck finding appropriate caveats. You may find a few in news stories based on the first release, because of the way it was framed.  Journals, and their news release writers, can do so much good, or not.