Science vs. Science “Journalism.”

Several decades ago, I was doing research trying to find drugs to prevent a certain tropical disease.  In this particular disease, the route of infection was through the skin, so we were looking at compounds that could be applied to the skin.   A couple of years in, we found one, which went by the identifier “593.”   It was quite simply the single most effective compound we’d ever seen.  It worked at very low concentrations, it had lasting power,  and it even survived repeated washing.   Success, right?  Not really.  It never made it past those early stages.   Our initial happiness with it turned out on further investigation to be premature, because the compound had a problem.  It caused “phototoxicity,” which means that it significantly increased your sensitivity to UV radiation.  A sunscreen with a negative SPF factor.  Which is not what you want for something you’re going to apply to the skin in a tropical environment.     Fortunately, we didn’t have news agencies reporting our initial results, because yesterday, I got an unfortunate chance to see see what happens when news organizations do just that – again.

What happened?  A group of pediatric researchers published a study looking at maternal “metabolic conditions” – that is, diabetes, obesity, and/or hypertension had any association with autism spectrum disorders, developmental delays, or other impairments.   The results?

Maternal MCs may be broadly associated with neurodevelopmental problems in children. With obesity rising steadily, these results appear to raise serious public health concerns.

In case you’re wondering, the “may” and “appear to” are standard conditionals in any scientific paper.  They’re there because of some very simple reasons.  First, this is the first study to look at it; second, it’s a fairly small population size; and third, they were looking at a specific set of factors.   Which means that further studies, ones that look at larger populations or other factors in addition to those may find out something different from what they found or something that was overlooked.   This is the way science works.  This is a “first report.”  Now it has to be repeated by someone else.

But you wouldn’t know that if you turned on the news yesterday morning.  The headlines? Obesity in Mothers Linked to Autism, Study saysAutism Linked to Obesity in Mothers During PregnancyStudy Warns Of Autism Risk for Children of Obese Mothers; Researchers Find Link Between Obese Mothers; Autism in Kids.  All breathlessly hyping the results, leading most people to belief that there’s an ironclad association.  Which isn’t what the researchers are saying, and in the more “staid” reports, after you get past the headline, you find that out:

“Although the results of this study suggest obesity is a risk factor for developmental problems in offspring, one cannot assume that developmental problems in the offspring are due to obesity, and many other factors may be involved or responsible,” Adesman said.


She also noted that while the research found an association between obesity and autism/developmental delays, it did not prove that being obese causes autism or other brain problems in the fetus. The link may be indirect.

“It may not be the obesity itself, but other things that lead to obesity, such as genetics, or lifestyle, or diet,” Krakowiak said.

What this study has done is open up additional avenues for investigation. But if you look at the way the news media has reported this, you’d be forgiven for assuming that obesity is a cause of autism.   It’s not, and that’s where science “journalism” has fallen down badly over the past years.  There’s very little understanding of science in it, and it’s geared to the sensational aspects, to pump “the story,” than to report the actual truth.  Along the way, it’s doing a lot of damage.

Oh, as a sequel to my opening.  Even though 593 was not the answer, it was the “discovery compound” for a very active class of compounds.  One of its close relatives, 34068, was almost as good, without the problems, and today is being used.  But we might not have found it if 593 hadn’t been so good, but “failed.”



Filed under Science

4 responses to “Science vs. Science “Journalism.”

  1. Vic78

    Is the whole mainstream media just a tabloid now? Do they really have to sensationalize everything? They’ve already dumbed down the discourse. What more do they want?

    • It’s a combination of seeking eyeballs – hence, the more sensational you can make it, the better – and the fact that most journalists, even on the science beat, really don’t have an understanding of science to begin with.

  2. Yeah – teaching a social science (political science) where studies are even more open to interpretation and multiple causality (it’s hard to run controlled experiments) I drill it in to students to always be skeptical when someone says “a study proved…”

    • One of the things we used to get hammered with a lot was “correlation does not equal causation.” 😆 One of the things about population studies like this is that they’ve found a correlation, not necessarily a causation. This is actually a very good study, but like all good initial studies, it raises many more questions for further investigation than it answers. The thing is that the news media wants to portray it as a “we found the cause of autism!” when the authors are very careful not to say anything like that.