We knew first that air pollution from tailpipe exhausts causes respiratory diseases like asthma. Then the epidemiologists added heart disease, and the mortality estimates went up. Then cancer, and they went up again. The latest addition is dementia. The death rates don’t go up, which is not for many of us a plus.
I have two sample studies for you. (H/T James Ayre at Cleantechnica, and a commenter there.)
The first is a very big cross-sectional study in Ontario. Hong Chen from the Ontario Public Health Office et al got access to the postcodes and health records of 2.2 million oldsters in Ontario. Using the postcodes, they worked out whether they lived within 50 metres of a major road in 2001, and whether by 2012 they had been diagnosed with dementia (n=243â€ˆ611), MS, or Parkinson’s. Takeaway:
Living close to heavy traffic was associated with a higher incidence of dementia, but not with Parkinson’s disease or multiple sclerosis.
The hazard ratio (similar to relative risk but not quite for stats nerds) was 1.07 overall, but rose to 1.12 for those who never moved or lived in a major city. Given the huge sample size, these results are quite strong (â€œp for trend = 0Â·0349â€) . The effect size is quite low, but could easily understate the true risk from traffic pollution because everybody is exposed to some extent. Incidentally, the negative findings for MS and Parkinson’s are valuable confirmation of the soundness of the study; statistical noise would show up everywhere.
The study nicely complements a Swedish study by Anna Oudin from UmeÃ¥ University et al, including Steven Nordin from the famous Karolinska. This was part of the longitudinal Betula study on aging and cognition, running since 1988, with an ultimate sample of 1806 (n for Alzheimer’s and vascular dementia = 302). They measured cognitive performance themselves, ruling out diagnostic variations, and were able to control for many other possible causal factors. Takeaway:
If the associations we observed are causal, then air pollution from traffic might be an important risk factor for vascular dementia and Alzheimerâ€™s disease.
They got much higher hazard ratios than the Canadians: 1.43, which rose to 1.71 when they excluded a sub-cohort added later to the study and only retested once after 5 years. The difference could be down to study design â€“ and the Swedish one looks more thorough -, or a lower pollution baseline, or to diesels, which are far more common in Europe than North America.
Oudin et al cite other studies pointing in the same direction if you are interested. For now, the general public like us can take it that there is a very strong statistical association, of uncertain magnitude but at all events quite high enough for a reasonable voter to take action. Dementia is common: the cumulative frequency was 11% in Ontario, 18% in UmeÃ¥. A 10% increase in risk would then mean 20,000 extra cases in Ontario alone over a decade. That’s a lot of preventable tragedy. The two study areas were in well-policed regions with high environmental standards and low average density. You would expect much higher pollution exposures and incidence in say inner-city Chicago of Glasgow. (Calling Harold.)
But where is the causal link? There is direct evidence of neurological damage from pollutants. The problem is that unlike heart, respiratory disease, and cancer, scientists are still pretty much in the dark about how exactly dementia arises. So there is not and cannot yet be a smoking gun. In theory, it could be something else in traffic than exhaust pollution: noise; or dust from road wear, tires, and brakes.
Let’s step back a little and think about policy. The question for Hizzoner the Mayor is whether, like Ken and Boris, to push hard for pollution cuts in the city, through comprehensive monitoring, low emission zones, purchases of electric buses, garbage trucks and police cars, regulations on taxis and delivery vans, light rail, and so on. What does the public health lady say? “Tailpipe air pollution causes cancer, heart disease, and respiratory disease for sure, and quite probably dementia if you survive those”.
The addition of dementia does not add much scientifically to the already overwhelming public health case. It does add to its political acceptability â€“ everybody is rightly scared of dementia, for themselves and those they care about, and so far it’s untreatable.
The residual doubts about the causation of dementia are secondary. They need to be handled carefully, though, so as not to give free ammunition to the unprincipled opposition.
- Don’t claim certainty when it isn’t there yet. The high probability is scary enough. Now address the long shots.
- If it turns out to be noise, the case for electrification is just as strong, as evs are much quieter â€“ some are even fitted with noise generators to warn pedestrians and cyclists.
- The same goes for brake dust, as evs normally have regenerative braking and use the brake pads less.
- Fine road dust is another possibility. Macro road damage from vehicles as almost entirely down to heavy vehicles; cars barely register. It’s the weight on the footprint that does it. The same should be true of finer particles (my non-expert inference}. The electric transition does not help here. (I will suggest to the UmeÃ¥ team that they check for this factor using bus routes as a proxy for heavy vehicle traffic).
- Tire dust is also dependent on weight, but possibly not to the same extent, as cars travel faster. Again, the electric transition makes no difference. Tire dust consists of fine hydrocarbon particles, rather like those found in diesel exhausts. A priori, it is most unlikely that dementia (and other disease) is only caused by one sort of fine particles and not the other. If it’s NOX, ozone or other gases, these are only in exhausts.
Conclusion: it is possible that dementia is promoted by two forms of pollution from vehicles, road and tire dust, that will not be cut by the electric transition. It is most improbable that they are the exclusive factors. This doubt will be cleared up by more research. Even if they were the sole factors causing dementia, the health case for electrification still holds up.
* * * * *
If it turns out to be dust, there is a technical fix to hand in the humble vacuum cleaner, inspired by this disconcerting canomorphic one.
The Canadian and Swedish studies were unsurprisingly in countries with working universal health care and good health records. The Canadian one in particular relied on number-crunching a centralized database of health records. I am fine with this, provided the data are anonymised â€“ not a problem with large datasets like these. If we hope to benefit from medical progress, the implied contract is that we are all experimental subjects. Crunch away.
Now where did I leave my glasses?
5 thoughts on “Gas-guzzlers cause dementia”
The Swedish study looks much more convincing to me than the Canadian one. As far as the Canadian one, I think I would put the bar much higher than p < 0.05 with that many people in the sample. That said, I find the effect plausible, and agree in any case that a push toward decarbonization of transportation would greatly improve public health, regardless of whether this particular effect exists or not.
I have red this blog post, but not (yet) the linked studies. However, as soon as I saw the post, a question leaped into my mind, in all capital letters:
WHAT ABOUT LEAD?
The age range of the folks currently and recently developing dementia seems to coincide with the folks who were subjected to the highest concentrations of lead in the vehicle exhausts of the fifties and sixties, when they were at the tender young age when their brains were most susceptible to lead poisoning.
The Swedish study is available online. I searched it for the word "lead" and found none. The Canadian study is gated, and the price for that one article is $31 U.S., which I find unacceptable to me. In the summary and in the abstract there is no mention of lead, and it would not surprise me to discover that it's never mentioned in that study either.
I thought of that. But one of the findings in the Canadian study is that the HR went up a lot (from 1.07 to 1.12) for people who did not move during the study period. Very few people spend their whole lives at the same address these days, so exposure to childhood lead would have been spread more or less randomly across the cohort. At all events, it was residence later in life close to (lead-free) traffic that made the difference.
Agreed on the price-gouging by the Lesser Vampire Squid, Elsevier. The Lancet should cut the tie. The research was paid for by Ontario taxpayers, why should they be overcharged again to see the results? I did not buy the full article and relied on the abstract plus Ayres' reporting.
Rather than zipcode proximity to a well-travelled road, wouldn't it be simpler to look at urban versus rural? I've spent my entire adult life in Cambridge, NYC, Boston, New Haven, Tokyo, and Queretaro, always within spitting distance of a traffic light, so idling cars have been my boon companions for 35 years. Plus, well, walking around, you are never more than five feet from cars whose engines are on. So shouldn't there be a more profound differential between urban and rural, triflingly easy to tease out of meta-studies on dementia?
You can find the authors' emails in the links if you want to ask them. I imagine they chose the harder route for a reason. One problem may be that practically nobody lives in the deep countryside: we are almost all urban or suburban. As you say, exposure to exhaust pollution is inevitable for everybody but Jeremiah Johnson, even for Amish. This means that the hazard ratios only give a pointer to the differential impact of high traffic pollution, and not the overall impact of all traffic pollution, almost certainly higher. The other thing is that medical concern focuses on those at high risk, and so should public policy.
Comments are closed.