Scientists Use The Double-Dog Epidemiologist Fallacy To Claim Breathing Induces Antibiotic Resistance

Thanks to AS! I don’t have your email, so I hope you see this.

The epidemiologist fallacy occurs when a scientist announces, directly or implicitly, that X Causes Y, but where X is never measured, and the “cause” is “confirmed” by his wee p-value.

It may be in the course of life that X really does cause Y. Even scientists get lucky. But to say that one knows this when X has never been measured, and no real measure of cause has been taken, and no causal mechanism known (rather than guessed at), is hubris times twelve.

This is the staple fallacy of epidemiology. Without this fallacy, the field would not exist as a going concern.

I once told this to an audience of academics, illustrating my thesis with prime examples of scientific goofiness caused by the epidemiologist fallacy. Some of them wept. Not for the mistakes and punishing over-certainty. For the reputation of Science, which they would not have disparaged in public. Rumor is some of those audience members are still in therapy because of this.

That effeminate sensitivity goes some way in explaining how this fallacy continues, even after its obvious flaws have been pointed out, again and again and again.

Our latest example is the peer-reviewed paper “Association between particulate matter (PM)2.5 air pollution and clinical antibiotic resistance: a global analysis” by Zhenchao Zhou and others in The Lancet: Public Health. You will recall that “PM2.5” is dust of a certain size.

Now saying—indirectly, with the wink-wink word association—dust is causing antibiotic resistance is a mighty claim. It must be backed up by muscular evidence.

Surely these scientists gathered a range of people, measured the amount of PM2.5 each had inhaled over some specific period, and then measured in these individuals whether or not each demonstrated resistance to antibiotics, taking great care to control (in the physical sense) for other possible causes, like previous antibiotic usage, type and dosage measured in each person.

For this is the only way one can have good evidence that PM2.5 is associated with antibiotic resistance.

Alas, our good authors did not do any of these things.

So how can they make the claim that X causes Y? Yes. Our old friend, the epidemiologist fallacy.

Before we come to the details, I am anxious to concede that the EF is wonderful in generating The Science—and headlines. Here’s one: “Air pollution linked to rise in antibiotic resistance that imperils human health“. Imperils human health!

Why, incidentally, did the propagandists feel it necessary to add “human”?

Now for the hilarious quotation from the propaganda piece:

The main drivers are still the misuse and overuse of antibiotics, which are used to treat infections. But the study suggests the problem is being worsened by rising levels of air pollution.

The study did not look at the science of why the two might be linked. Evidence suggests that particulate matter PM2.5 can contain antibiotic-resistant bacteria and resistance genes, which may be transferred between environments and inhaled directly by humans, the authors said.

Read that last paragraph again. Did you see? The authors pronounce the wild guess that dust contains antibiotic-resistant bacteria and resistance genes!

It would take the heart of an Expert not to laugh at that.

Here’s what the authors did. It was not only the epidemiologist fallacy, it was the double-dog epidemiologist fallacy. This is when in “X causes Y”, not only is X not measured, neither is Y!

My emphasis:

…we created a dataset of antibiotic resistance patterns across 116 countries using data from 2000 to 2018 from the ResistanceMap…

Data on antibiotic use (ie, defined daily doses [DDD] per 1000 people per day) had been collected from 2000 to 2018 for 116 countries in a previous spatial modelling study…[PM2.5 was] collected from the air-quality databases of WHO, the Health Effects Institute, and the European Environment Agency for 2000–18 for 116 countries.

So no X measured and no Y measured on any individual. And antibiotic use was not only not measured, it was the output of another model!

Then came their own statistical models, the details of which are not of interest today. Basically per-country (not individuals!) regression, which is often used to fallaciously claim causation.

Obviously, more people live in cities than before, there is more dust in cities, and antibiotic resistance is increasing because of ask-your-doctor-if-this-drug-is-right-for-you propaganda, so we’d expect at least some vague correlation with being in or near a city and resistance. And that’s what they see.

Yet the began their Discussion with these words: “This analysis used data collected from worldwide surveillance reports to estimate the effect that PM2·5 has on antibiotic resistance.”

My friends, this is causal language. Strong words. Words not backed up by their statistical manipulations.

But I hope you can now see why I call it the epidemiologist fallacy.

Subscribe or donate to support this site and its wholly independent host using credit card click here. Or use the paid subscription at Substack. Cash App: $WilliamMBriggs. For Zelle, use my email:, and please include yours so I know who to thank.

Categories: Statistics

7 replies »

  1. Briggs you’re getting off into the weeds on this one the only criteria
    of success today of any scientific theorem is how many inches of newsprint
    it can fill and how much fear and anxiety it can inculcate in the general population.
    By this measure though this study will never be up there with cow farts it is a resounding
    success worthy of increased government funding. What the government really needs
    is something like a fear-o-meter to correlate suicide rates, despondency, & anhedonia, with
    psychiatric admissions to better target taxpayer funding for more newsworthy research like

  2. So, breathing in causes antibiotic resistance, and, of course, breathing out expels the deadly pollutant CO2 (Carbon!!!). Experts caution everyone to stop breathing.

  3. I believe laziness (aka, “efficiency”) explains a lot of curious space-filling frivolity produced these days, whether it’s done in the service of current news sourcing or the structuring of scientific studies. Why risk physical exertion researching new content when you can sit at your computer and mine existing content from a seemingly infinite online supply? Apparently this now also extends to the choice of the actual subject matter of studies. I’d bet real money that there exists an academic version of that famous spurious correlations website (with more obscure source data, the most obvious humor removed, and a large dash of political correctness added) that is used to come up with these strange studies.

    The flip side of this is that the study authors don’t have to worry about reproducibility – who’s going to bother? Bad studies are like those lies that make it half-way around the world without any pants on. Or something.

  4. It’s models on top of models within models riding atop models all the way down, and none of them are as sexy as they seem whenever you take a closer look. THE SCIENCE ™ must therefore be drunk given that X here always immeasurably leads to Y and always wakes up in a mistaken mess!

  5. Over the last 3 1/2 years, I made a pretty big list of people I hope to save by stopping their breathing.

  6. Thanks for this. Today I dared to turn on the morning news shows as an experiment to satisfy my curiosity and observe the trend in health education by C?A agents posing as News Anchors posing as unqualified health educators. I’m a qualified health educator. I studied statistics 2 terms. Well, you may need to be aware that the trend this week is COVID. It’s baaaack. Actually, I’m two steps ahead. As Lenin said, “One step ahead, two steps back.” What I really observed on the news/gossip shows was a string of mind numbing pharma ads, warning of rare diseases, and the like. TV News: don’t go there alone.

  7. Asked Alexa what the epidemiologist fallacy is:
    She said,
    “In brief, the epidemiologist fallacy is this: When an epidemiologist says, or hints, with a wink in his eye, that X causes Y but where X is never measured and where the cause is proved with wee p values”!

    Who is she quoting, I wonder?

Leave a Reply

Your email address will not be published. Required fields are marked *