Following the publication of a scientific paper analysing heart condition and the impact of air pollution, the national media immediately jumped to headlines such as “Low Levels of Air Pollution Linked to Changes in the Heart” as published by the BBC. Stories were typically illustrated with pictures of traffic jams, car exhaust and visible London air pollution. Calls to reduce legal maximum air pollution figures were added.
What does this scientific paper actually show? You can read it here: https://www.ahajournals.org/doi/pdf/10.1161/CIRCULATIONAHA.118.034856. It’s a paper by Nay Aung et al. But here are some comments on it:
- The paper was published in the journal “Circulation”, a small distribution specialist scientific journal, It is based on information from the UK Biobank which contains medical information on thousands of volunteers who have in this case had heart scans.
- The study correlated the information from 3,920 individuals to air pollution data (specifically NO2 and PM2.5) where they lived from 2005-2010.
- The study concluded that after adjusting for numerous other factors such as demographics (age, sex, ethnicity), anthropometrics (height, body mass), socioeconomic factors (income, employment, educational status), cardiac risk factors and physical activity there was a positive correlation between minor changes to left ventricle (LV) mass and other heart changes to air pollution levels where the volunteers were resident.
- Such changes to a heart are known to precede heart disease.
- The paper’s authors therefore concluded that “our findings add to the growing evidence of the damaging effects of ambient air pollution even in the setting of relatively low exposure levels”.
Are their conclusions justified beyond any doubt, and in particular are the popular media’s headlines justified? My conclusion is no. For example, like any epidemiological study it only provides a possible statistical association, not a direct cause. As Prof Kevin McConway commented on this paper (as reported by the BBC): “Heart disease is affected by a wide range of factors – smoking, drinking alcohol, diet, exercise, social position, and more. Suppose that people whose heart health is worse because of some of these factors are also more likely to live in places where air pollution is high. That could show up as a correlation between air pollution and heart disease, even if the pollution itself is having no direct effect on the heart”.
Another possible issue is that air pollution inside houses is known to often be many times worse than that in the most polluted streets. That pollution comes from cooking, new paint, fabrics, carpets, smoking by other residents, animal hair, etc. Lack of ventilation in houses and apartments can increase levels substantially so people who live close to noisy roads who never open their windows as a result may be particularly affected.
The report is open to attack on the detail of their statistical methods, and they also note that other similar studies did not provide the same evidence in all cases. In summary the overall evidence is quite weak. Neither does the report confirm that the minor changes noted to heart mass lead in this case to significant heart disease.
Their reference to “low exposure levels” may also be misleading because air pollution levels were not measured outside the volunteers’ houses or where they work. In addition the fact that the people studied were volunteers, i.e. were self-selected rather than being a randomised sample, could have biased the outcome even though lots of adjustments were made for possible confounding factors.
All the report really suggests is that more study should be undertaken of a possible effect. The conclusion drawn by some commentators that air pollution legal limits need to be reduced further is not substantiated by this report.
In the meantime, readers are advised not to live within a few metres of a busy road because it may be bad for your health. But that’s no surprise is it? Just living in a noisy environment is known to be very damaging to your health. High noise levels are correlated with cardiovascular disease according to the World Health Organisation. It seems it increases stress levels which has a negative impact on health.
What the Aung report does not do is justify even more aggressive attempts to reduce air pollution in cities such as London, where NOX and particulates are already falling after the mistaken support of more diesel vehicles by the Government. Road vehicles will soon no longer be a major contributor to air pollution in cities so more scaremongering of this ilk is not required.
The answer to the question posed in the headline of this article (“Do Low Levels of Air Pollution Damage Your Heart”) is simply that it is “Not proven”.
There are four ways a study like this can give misleading answers.
1 The sample. The authors do not pretend the sample is representative of the general population, and they rightly say that this doesn’t matter for this kind of comparative study. They have gone to great lengths to remove factors which might affect the outcome, like smoking, but they may well have missed something which means that those predisposed to heart problems for reasons unrelated to air pollution may tend to live closer to roads.
2 Errors in the methodology. Hard to evaluate when the data and methods are not present, but I’ve encountered situations before when corrected datasets from different sources and analysed by different people actually cancel each other out and produce a nice relationship between the two factors where none exists. A classic case of this was the TFL report linking average speeds to accidents on rural roads, where the speeds and the accident numbers had both been adjusted for traffic levels.
3 Other environmental factors are causing the heart problems and these are related to where people live. The types of housing near to roads tend to be quite specific due to 1950s ribbon development, and its entirely possible that these houses have different issues from newer houses on estates.
4 There is a problem with air quality but its the internal air quality in the houses, generally far worse, causing the issue not the external levels of pollutants. It is quite likely that the further you are from a road the less likely you are to have gas cookers, for example. Or that the levels of pollutants you are exposed to are very dependant on the size of layout of your home, and this is linked to distance from main roads. Hard to say without any data, but a UK study of internal NO2 levels showed a link with childhood asthma incidence but one in the Netherlands did not – a clear indication that other factors are at work in the UK.