Air pollution directly linked to increased dementia risk

This article repeats the common mistake of conflating correlations and causality. The main results are (1) that PM2.5 exposure is correlated with dementia in humans, (2) some experimental results with mice. This does not establish causality in humans. The paper is careful to stay juuuust on the right side of the line by carefully using "associated" in the right places. But the press release discards that pretense at rigor and jumps straight to full-on claims of causality in people:

> Long-term exposure accelerates the development of Lewy body dementia and Parkinson’s disease with dementia in people who are predisposed to the conditions.

I think it's entirely possible (perhaps even likely) that this is true. But the paper does not show it.

Why is this brought up here every time an association article is mentioned, every undergraduate that took a statistics course has covered the difference between correlation and causation.

We do population level correlation studies because sometimes a double blind study is unethical, and double blind studies is the bar for establishing causation in the medical community. We cannot give one person randomly worse air to breath, and even if it were experimentally feasible it would be ethically impossible because there is strong suspicion we would be harming the subjects.

Let’s discuss the actual data. The dose dependent result that was found is an indicator of a strong relationship. There is a clear potential mechanism of action (Air -> Lungs -> Bloodstream -> Brain). This isn’t a controversial result in the literature, it’s more evidence for what we already know —- air pollution is very bad.

The maps in the article wasn’t even super clear when you inspect it visually. Didn’t read the study but it can’t be that strongly correlated.

You seem to be responding as if I claimed that population-level correlational studies are bad, or that I claimed that air pollution is not bad for you. But I did not claim either of those things.

What I claimed is that this press release takes a population-level correlational study and presents it in a misleading way that implies causation was established. Which this press release most certainly does.

I was going to say something similar. Obviously I was just taking a visual look and not doing any rigorous analysis, but California certainly sticks out like a sore thumb. It has many areas in the Central Valley and Southern California with very high PM2.5 levels, but no "purple" areas with high dementia levels in the state. I also just read the article and not the study but I'd hope they give an explanation for that glaring outlier.

I'd be interested to compare the disease map with a map of average income, because at first glance the disease data looks to be correlated with wealth, and we already have tons of research that shows that wealth is one of the biggest determinants of health outcomes in the US.

Oh yeah, sure definitely it's just as likely that people predisposed to dementia move to places with high air pollution...

Another Red state theory about Blue cities ?

It is unfortunate that the title and subheading of the article use a causal phrasing. Fortunately the body of the article does maintain the correct distinction.

I noticed this too. I do wonder what it would look like if you controlled for relocation (moving to another region and then developing symptoms), disease onset (both maps are for the same time ranges), and the types of pm 2.5 exposure.

The correlation != causation argument at large is usually a sophomoric dismal of valuable data, and I’d rather see discussion on the data or specifics to research methodology then this meta criticism.

Ultimately both exaggerating pop-sci articles and dismissive comments are contributing to public distrust of science.

Probably not that, but it could be something else that also correlates with pollution like living in big cities, or working in a factory, walking more/less, noise levels, lifestyle, etc

Maybe the thought-terminating cliche should be posted as a comment by a bot at the top of every submission so there's a home for this sort of complaint.

Another fossil fuel shill

Well, in some cases we know that they do.

Dementia is linked to diabetes. And diabetes risk is increased for African-Americans. And African-Americans live in high-pollution urban areas for entirely historical reasons.

So some amount of the causation here does go in the way opposite to what a person might naively suspect.

> Dementia is linked to diabetes. And diabetes risk is increased for African-Americans. And African-Americans live in high-pollution urban areas for entirely historical reasons

A is correlated with B. B is causally correlated with C, i.e. C causes B. (C is correlated with D.) Hence C causes A.

Let’s replace. Flowers are correlated with bees. Bees are caused by hives. (Hives are correlated with trees.) Hence, hives cause flowers.

Loosely, yes. Formally, no.

There might be some hidden variable at play that's correlated both air pollution that is really causing the dementia. Like street noise. Not saying it's likely, but it's not impossible.

It's even worse... dementia sufferers are the ones emitting these particles ;-)

We know that diabetes causes some amount of dementia and that flowers cause no amount of bees. And so on. Your example is specious, and obviously so.

> This isn’t a controversial result in the literature, it’s more evidence for what we already know —- air pollution is very bad.

It is certainly controversial for individuals whose way of life (see non-electric car-centric society) is being questioned by this science. Just look at some of the cynic answers in this thread pretending that air pollution is not bad

Thanks, got a good laugh out of this one.

You’re being sarcastically dismissive, but this is a real possibility. The real world is complicated and disorders with environmental effects are often multi-factorial.

Air pollution might not be the direct cause, it might be a proxy measurement that is correlated with some other factor or factors that contribute to dementia risk. For example, do areas with higher air pollution measurements also have higher or lower rates of something else that is actually contributing to the dementia directly? Do they simply correlate with overall development of the area, and therefore areas with poor pollution numbers also have high levels of water pollution?

People who are poorer and have worse health are predisposed to live in cheaper dwellings, many of which are closer to roads and thus more noisy and with more air pollution.

People who are poorer and have worse health, also have an increased incidence of dementia, seemingly independently of the number of particles in their dwellings.

Outlive (book) talks extensively about dementia risk and Alzheimer's as "type 3 diabetes".

New Orleans is another one.

I’m not sure this is a completely valid statement. Take Ontario - the most expensive places to live are in Toronto with the most traffic, the cheapest places are more rural without that heavy traffic and thus less pollution.

The Parkinson's Plan talks extensively about the risk factors associated with the disease, and discussed the link between air pollution and the disease. They say that it is a trigger / cause that only sometimes works for reasons unknown (paraphrasing). Is a "sometimes cause" still causal?

In rural areas it's pesticides. And in urban areas it's dry cleaners and air pollution. Parkinson's Plan is worth a read for the kind of details you can't get on a single article

(OP makes a good point, just going on a slight tangent here)

We really need a term that sits between correlation and causation in situations where data is difficult to come by. There's such a huge rift of meaning between these terms, and too often 'correlation is not causation' gets wheeled out in a room of people that already know that and are trying to figure out the nuances.

How about plausal? Aka it's rather plausible that there is a causal relationship between two things but causality is hard to prove.

"Air quality and dementia have a plausal relationship".

The bar for plausation is much lower, yet many correlates still won't meet it. "Bad air quality causes dementia" is a categorically different statement than "ice cream sales cause shark attacks", if we establish the category of plausal relationships.

Just because air pollution is bad, doesn’t mean it is bad in this way. It could just as easily be that some other lifestyle thing is causing dementia AND leading to living in areas with worse pm2.5.

Especially since pm2.5 tends to be higher in areas with higher population densities, near roads, industrial areas, etc.

Hell, maybe the underlying major risk factor is actually time spent in the car. Or repeated viral exposures of a specific type. Or a specific type of air pollution.

Either way, it’s not like everyone is going to be moving out of high pm2.5 areas anytime soon, or that we’ll be able to just solve the sources of pollution right now even if it is the cause.

How is pm2.5 not something we can have a direct impact on based on environmental laws?

Ask California and all the wildfires.

How is wildfire frequency and severity not directly impacted by regulation?

Wildfires are already illegal in California. As is producing a lot of noxious smoke! They even have an entire (rather large) government division (CALFIRE) responsible for stopping them.

For some reason, nature DGAF.

Or do you have some other policy proposal? Banning lightning or fire perhaps? Making trees illegal?