Wednesday, 18 February 2026

Spurious correlation news

The "gateway" theory of vaping is based on the observation that teenagers who have ever vaped are more likely to smoke than teenagers who have never vaped. I have often pointed out that teenagers who have ever vaped are more likely to have done lots of things that never-vaping teenagers have done, because they are different people. That doesn't mean that one behaviour causes another.

In 2018, for example, I wrote...
 

Since marijuana has never killed anyone, supporters of the war on drugs resort to claims about it being a gateway to heroin. E-cigarettes have never killed anyone either and so anti-nicotine extremists resort to claims about vaping leading to smoking.

In both cases, correlations can be found, but it is fairly obvious that they are not causal. Teenagers who vape are more likely to smoke, but they are also more likely to ride motorcycles, watch X-rated movies and have unsafe sex. They are also more likely to smoke cannabis, for that matter, but that doesn’t mean that vaping leads to any of these behaviours, nor would they be less likely to engage in them if vaping didn’t exist.

 
In an IEA report the previous year, I had written... 
 
Indeed, one of the studies in the Soneji review found that e-cigarette users were not only more likely to smoke cigarettes but were more likely to smoke marijuana (Unger et al. 2016). It would not be surprising to find that they are also more likely to drink alcohol and have unprotected sex, but it would be a stretch to claim that these risky activities are somehow caused by their earlier experiments with vaping.
 
And only last month I made the same point in relation to zero-alcohol beer...
 
Teenagers who like the idea of drinking non-alcoholic beer are presumably more likely to be interested in drinking real beer in the same way that risk-taking teenagers who are drawn to vaping, motorcycling and unprotected sex are more likely to be interested in smoking and illegal drugs. This is known as a ‘common liability’ and it could produce a statistical correlation between non-alcoholic beer consumption and actual beer consumption, but it would not be serious evidence of cause and effect, i.e. a ‘gateway’.
 
The example of vaping as a gateway to unprotected sex is a good way of illustrating spurious correlations. There is no plausible mechanism for vaping - or any form of nicotine use - to cause people to have unsafe sex. 
 
So it was with grim amusement that I saw that some clown in (where else?) Australia takes these correlations seriously.
 

Emergency medicine expert Professor Brian Burns said the anti-vaping program was vital to help curb vaping among youth given it was a gateway to cigarette smoking and high-risk behaviours. 

“Studies have shown that engaging in unsafe sex, other substance abuse, drink driving, texting while driving and driving without a seatbelt are associated with increased e-cigarette use among youth,” he said.

“Sensation seeking – the desire to experience novel sensations and the willingness to take risks is also associated with e-cigarette use. These activities can result in severe physical injury and harm.”

 
This chump also claims that vaping causes "asthma attacks", "inflamed or collapsed lungs", "severe respiratory illness", "seizures" and "cardiovascular system shutdown". Australia really is a basket case.
 
The article is headlined 'World-first anti-vaping program rolled out to tackle teen nicotine addiction crisis'. Perhaps they should try banning e-cigarettes, lol.


No comments: