August 20, 2019

expert reaction to study looking at air pollution and age-related macular degeneration

A study, published in the Journal of Investigative Medicine, reports a link between long term exposure to pollutants from vehicle exhaust and a heightened risk of the common eye condition age-related macular degeneration (AMD).

Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“Age-related macular degeneration (AMD) has been estimated to affect well over half a million people in the UK, to the extent of causing some loss of vision, and there are about 70,000 new cases a year, the majority being in elderly people in their 70s or 80s. Its causes are not entirely understood, so research on potential risk factors is welcome. However, this new study leaves a lot of questions to be answered.

“I think this study provides reasonable evidence that high levels of the two pollutants covered, nitrogen dioxide and carbon monoxide, could well be a risk factor for AMD. But it’s far from certain that these two pollutants definitely do increase the risk, and if so, by how much they increase it. Maybe the results can be explained by other factors, or by problems in measuring pollutant levels. Even if air pollution is the real cause of the increased risk, it could be that the method by which the pollution increases the risk is due to some other pollutant that also reaches high levels when nitrogen dioxide or carbon monoxide is high. High levels of different traffic-related pollutants do tend to occur together. The researchers call for more research, that could tie down more clearly whether these pollutants are risk factors, the extent to which they increase risk, and how they increase risk (if in fact they do). I’d support that call. But I think that there are many more important reasons for concern about air pollution than our current knowledge of its possible association with AMD.

“This is an observational study, and it is always difficult or impossible to establish what causes what in such studies. The results make it clear that there are several large differences between people living in areas with different levels of the two pollutants that were considered. Unsurprisingly, the areas with the highest pollution levels were more likely to be urban, but there were also other differences, for instance, in the patterns of other disease that the participants had in areas with different pollution levels.  Such other differences may explain some or all of the differences in rates of AMD diagnosis that were observed, rather than the different levels of air pollution. The researchers made some statistical adjustments to allow for some differences, but they could not make adjustments for whether the participants smoked, or their obesity level, or their family history of AMD, because they had no data on these things. But smoking, obesity and family history of AMD are known to be risk factors for AMD, and they might differ considerably between areas of low and high air pollution. So, the study can’t tell us to what extent the differences in AMD rates might be explained by differences in these factors rather than in the levels of nitrogen dioxide or carbon monoxide.

“The researchers also had some problems in estimating the levels of air pollution that the participants were subject to. The research used databases from which people’s addresses had been deleted, for reasons of privacy. So, the researchers could not use the measured level of air pollution at the participant’s home, because they did not know it. Instead they used the location of the health care facility where the participants had got treatment for an acute upper respiratory infection – that is, a cold or sore throat or similar illness. This leads to three problems. First, the treatment location would not usually be exactly where the participant actually lived, and the pollution level at their home could possibly have been quite a lot different. Second, people who had not sought treatment for a cold or similar ailment during the study could not be included at all, so that the participants in the study might not have been a very representative sample of the whole population aged over 50. Third, no account could be taken of air pollution levels inside the home, or at work, or in other places where the participants spent a lot of time. This third problem is quite common in studies of potential effects of pollution, but the first two are not, and may have led to considerable inaccuracy in estimation of air pollution levels.”

Prof Chris Inglehearn, Professor of Molecular Ophthalmology, University of Leeds, said:

“This is an interesting study but one that needs to be interpreted with caution, as the authors themselves state. What they have shown is that ARMD tends to be more common in geographic locations that have high air pollution. However, there are many other disease risk factors that come with life in such places, and correlation does not prove causation. The study has taken account of socio-economic status and, indirectly, has allowed for smoking, a major known ARMD risk factor, but no data exists for, for example, diet or genetic risk. A contribution to ARMD risk from air pollution is plausible and the findings of the study do suggest that may be the case, but a doubling of frequency due to air pollution alone is harder to believe. It seems more likely that air quality acts in combination with other geographically correlated risk factors.”

‘Traffic-related air pollutants increase the risk for age-related macular degeneration’ by Chang et al. will be published in Journal of Investigative Medicine at 23:30 UK time on Tuesday 20th August, which is also when the embargo will lift. 

DOI: 1136/jim-2019-001007

Declared interests

Prof Kevin McConway: “Prof McConway is a member of the SMC Advisory Committee, but his quote above is in his capacity as a professional statistician.”

Prof Chris Inglehearn: None