Article Text


Out of Africa
  1. B Brunekreef
  1. Correspondence to:
 Prof. B Brunekreef
 Institute for Risk Assessment Sciences, Utrecht University, PO Box 80176, 3508 TD Utrecht, Netherlands;

Statistics from

Commentary on the paper by Venn et al (see page 376)

Many studies from the west have implicated traffic pollution in respiratory disease, especially in children. Usually, these studies have been conducted in areas with high traffic densities such as inner cities, or areas near major highways. The daily number of vehicles passing on roads near residences or schools of study children have typically been in the tens of thousands, sometimes well over one hundred thousand. Air pollution studies conducted near such roads have shown relatively high concentrations of traffic related air pollutants, compared to sites away from the direct influence of busy roads. Typically, such contrasts were markedly larger for specific components such as NO2, soot, PAHs, and benzene than for particulate matter metrics of current regulatory interest, PM10 and PM2.5.1,2

In this issue, a study is presented from Jimma, Ethiopia. Jimma is a small town of some 100 000 inhabitants, located far away from the main population centre in Ethiopia, Addis Ababa.3 It has no major industry, few inhabitants can afford cars, and one would expect air pollution levels from traffic to be exceedingly low. Yet this new study suggests that subjects living close to the small number of paved roads that Jimma has, experience more wheeze, the closer they live to these roads. However, median traffic density during the daytime was only 653 vehicles, with a range of only up to about 2500 vehicles per 12 daytime hours. This is some two orders of magnitude less than the traffic densities studied in the west, and it is puzzling that relatively similar findings were obtained.

So how likely is it that 600 vehicles per day produce enough air pollution to adversely affect the health of subjects living close to the road? The answer perhaps is in the age and maintenance of the vehicle fleet. We sometimes forget that in developed countries, cars have become incredibly cleaner in a period of decades. This is due to changes in engine design, to changes in fuel composition, and introduction of catalytic converters. Also, the introduction of mandatory inspection programmes for older vehicles, including emissions testing, has likely resulted in the timely removal of the dirtiest part of the vehicle fleet from our roads. Ironically, whereas some of these vehicles no doubt end up in our own scrap yards, others start a second, most likely even more polluting life in poor countries without the proper institutions or funds to do much of anything to reduce pollution. Not too many references can be found to quantitatively document what the results could be. A report from Tehran, Iran4 argues that the 2 million vehicles on the roads are so old that many are highly polluting. Surely, the PM10 concentrations reported for Tehran are in the hundreds of μg/m3, much higher than anything found in similarly sized cities in the developed world with even larger vehicle fleets. But even in the developed world, despite all our efforts, so-called “super emitters” are still said to contribute significantly to air pollution concentrations, in the order of 50% of emissions being generated by just 10% of the vehicle fleet.5,6

So imagine what it would be like to have a vehicle fleet consisting largely or entirely of “super emitters”. A picture tells a thousand words, and fig 1, kindly supplied by Dr Venn, illustrates the point forcefully. It seems at least conceivable that 600 vehicles in Jimma count for a lot more, in terms of pollutant emissions, than 600 vehicles in Nottingham. The fact that Jimma lies at 5500 ft elevation adds to fuel consumption, and therefore to pollution as well.

Figure 1

 A traffic scene from Jimma, Ethiopia.

Some others aspects were noticeable. The investigators made separate counts of diesel vehicles, but could not show that these were more specifically related to wheeze than total counts, as in our earlier work in the Netherlands.7,8 However, Dr Venn told me that the correlation between diesel and total vehicle counts was high (as is often the case), so that the study was not in a good position to actually make the separation. Another is that vehicle counts were made several years after the symptom questionnaire was administered. This could be a problem when important changes over time in paved road network and/or road usage would have occurred, but this does not seem to have happened (Dr Venn, personal communication).

Venn and colleagues’ paper also shows that the prevalence of wheeze was not smaller in subjects living away from the roads than in the subgroup living within 150 m, in which a distance-response relation was found. This shows, as one would expect in a multifactorial condition such as wheeze, that overall, other factors than traffic fumes are likely to be more important as determinants of symptom prevalence. Nevertheless, the study shows that at least in some part of the population, symptoms were caused, or exacerbated by traffic fumes. Welcome, Jimma residents, to the developed world ...

Commentary on the paper by Venn et al (see page 376)


View Abstract


  • Competing interests: none declared

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles