There is convincing scientific evidence that exposure to air pollution, in particular ambient particulate matter (PM), is related to both short and long-term health effects. Increased mortality and hospitalizations for cardiopulmonary causes have been noted in several studies evaluating the short-term effects of PM₁₀ or PM_2.5[1]. Some longitudinal studies conducted in the US and in Europe did find a consistent association between long term exposure to air pollution and natural mortality, especially for cardiovascular diseases [2–7].

In urban areas a relevant source of air pollution is vehicular traffic as the importance of other sources (e.g. industry, power plants) is declining. In some of the studies evaluating health effects, proxy measures of exposure have been used, such as distance from busy roads, distance-weighted traffic density or the length of main street segments within a buffer from home [8–12]. These proxy measures tend to be correlated with other more objective measures (for example, NO₂ concentrations estimated by dispersion modelling and/or land use regression models) and are relatively easy to obtain using Geographic Information System (GIS) [13]. Some studies have directly assessed the relationship between living along busy roads and mortality or morbidity. For example, Hoek and colleagues found, in a cohort of adults aged 55-69 years, that cardiopulmonary mortality was associated with living near a major road [5]. In the SAPALDIA (Swiss Cohort Study on Air Pollution and Lung Diseases in Adults) study an increased risk of regular phlegm and wheezing in non smokers living within 20 m of a main street was found [8]. All the literature has been recently reviewed by a scientific panel from the Health Effects Institute (HEI panel on the Health Effects of Traffic-Related Air Pollution) in the US [14]. The Panel concluded that the evidence was "sufficient" to infer a causal relationship between exposure to traffic and exacerbation of asthma and "suggestive but not sufficient" to infer a causal relationship with onset of childhood asthma, non-asthma respiratory symptoms, impaired lung function, and total and cardiovascular mortality.

When evaluating the relationship between living along busy roads and mortality/morbidity, the potential role of socioeconomic status should be considered. The relationship between socioeconomic position and exposure to air pollution has been extensively reviewed and "environmental justice" concerns have been raised [15–21]. In fact, low socioeconomic groups of the population seem to suffer from the worst environmental conditions, including poor air quality [18–24], and tend to be more vulnerable to air pollution [19, 25–27]. However, the extent to which low socioeconomic position and proximity to busy roads are related is not well-defined.

Few studies have characterised subjects living near high traffic roads in urban areas to better understand the exposure characteristics and the susceptibility factors of air pollution. The lack of studies is somewhat surprising given the potential public heath impact of traffic exposure. We established a large cohort of people who were resident in Rome in 2001 with information on socio-demographic characteristics and baseline health status to study long-term effects of traffic-related air pollution. We wished to use diverse indices of socioeconomic position (both at the individual and neighbourhood level) and different geographic information system (GIS) measures to characterize the residents who live close to busy roads in Rome. In the present paper we describe the main baseline characteristics of the cohort and examine the association of exposure to traffic with socio-demographic characteristics.

The total population of Rome in October 2001 was 2.56 million (2001 Census) people. For 2,118,670 residents (84%) we had information on the address of residence and on individual socioeconomic position. We selected only those individuals whose address was matched and geocoded by the Tele Atlas system with optimal quality, resulting in 1,898,898 subjects that were considered in this study. There were no significant differences between subjects included and excluded from the analysis for age, and gender.

The study clearly shows that a large fraction of the Rome population is exposed to traffic at home. This is not surprising when one considers that out of a population of 2.6 million inhabitants, almost 2.3 million vehicles were circulating in the city in 2001, including moped/motorcycles, and commercial vehicles. The public health problem of traffic is amplified as one considers that in 2001 a large proportion of the vehicles were in the high emissions categories (EURO 0 and EURO 1) and that 17.4% of them were diesel-powered vehicles. Exposure to traffic is not evenly distributed in the city: we found that all the measures of traffic exposure were higher in the elderly population than in younger groups. In general, traffic was not higher in the poorest sector of the population, on the contrary, low SEP was associated with greater distance from HTRs and with lower traffic density within 150 m from residence. However, when we restricted the analysis to the city centre, higher exposures to traffic were found among subjects in the low socioeconomic groups. In any case it appeared that area level more than individual level drives the association between SEP and traffic exposure.

The explanation for the traffic exposure and age and SEP relationships is clearly related to the urbanization history in Rome with the consequence that people of medium-high social class are located in the more central and prestigious areas of the city where there are more high traffic roads. The prices of dwellings in the city centre are much higher than in the suburbs and the poor air quality is rewarded by prestigious buildings, good quality of public transportation, and access to several facilities. This clearly explains why elderly are more exposed to traffic air pollution: in the past it was still affordable for young couples to buy a flat in the city centre, but now the value of dwellings is so high compared to the suburbs that in the last decades young new families have tended to live in the periphery. The same reasons explain the different socioeconomic distribution of population by area of the city, new residents in the city centre tend to be very well-off whereas less advantaged people move from the centre to the suburbs for economic reasons. It is interesting, however, that restricting the analysis to the city centre, the relationship between SEP and traffic exposure changes, with low exposures in residents better educated or who live in a high SEP area, suggesting that environmental inequalities do exist in the central area.

As indicated before, the large number of people residing close to busy roads in Rome underlines the need for a public health attention to the issue. Unfortunately, there are few comparable measures from other cities in the world to evaluate whether other places have similar distribution values to what we found in Rome. In the Netherland cohort study, almost 5 percent of the study members lived within 50 meters of a busy road, whereas it has been reported that the proportion of the population living within 50 meters from a major road in Los Angeles and Toronto is approximately 32% [6, 14]. A meta-analysis of the available evidence indicates that for most pollutants there is a distant decay gradient in the range of 100-400 meters from the source and therefore the population with direct exposure to traffic-related pollution is even higher [33].

Our results on the association between SEP and traffic exposure were similar to those from a study in the Netherlands [5], but in contrast with the results from studies on environmental equity from other countries, especially the US [15, 34–36]. However a study by Buzzelli and Jerrett found that in Toronto racial minority groups were less exposed to air pollution than other groups, and that dwelling values predicted total suspended particulates levels in Canada [37].

Most of the early evidence that people of lower socioeconomic status are more exposed to air pollution comes from studies that have considered point sources, mainly industrial settlements [14]. On the other hand, the recent literature available on exposure to traffic is more limited. Comparability is also difficult given the use of diverse measures of exposure to air pollution, and different measures of socioeconomic position. In the studies conducted in the US, it is taken for granted that disadvantaged groups of the population live in the worst environmental conditions including high exposure to outdoor air pollution [17]. It should be recognized that the literature on the issue is still limited and additional observations are needed from different parts of the world. Our finding in Rome is not new. In a previous investigation on short-term effects of PM₁₀ in the city, we pointed out that emissions of PM, CO, NOx, and benzene were higher in areas of higher socioeconomic status [25]. In addition, in a large cross-sectional study of adults, GIS indices of exposure to traffic and estimated NO₂ level at the residential address from a land use regression model were directly linked with socioeconomic position [13].

Despite the socioeconomic gradient of exposure, the short-term mortality effects of PM₁₀ in Rome were larger in the low socioeconomic groups perhaps as a result of a different susceptibility level of poor people rather than to different proximity to traffic [25]. On the other hand, the fact that higher socioeconomic groups live in high traffic areas does not necessarily mean that they are more exposed than underprivileged groups are. More advantaged people work in air-filtered office buildings than outside, they use private transportation more than public transportation, they are more likely to live closer to their work places, they are more likely to spend the week-ends outside the city than low socioeconomic groups. Clearly, investigations on personal exposure to air pollutants are needed to better qualify the issue.

Important new evidence has been accumulating on the effects of long-term exposure to traffic-related air pollutants. Although the relationship between traffic-related air pollutants and cardiovascular mortality and morbidity has not been clearly defined yet, it is clear that particulate matter has an effect on the cardiorespiratory system [14, 38]. An important study on long-term effects of PM_2.5 on the cardiovascular system was conducted in the US [39]. The follow-up included more than 65,000 postmenopausal women (Women Health Initiative) without previous cardiovascular disease in 36 US metropolitan areas from 1994 to 1998. Each 10 μg/m³ increase in PM_2.5 was associated with a 24% increase in the risk of a cardiovascular event and a 76% increase in the risk of death from cardiovascular disease. A report from Oslo has linked long-term exposure to traffic-related pollutants to cause-specific mortality [40]. A consistent effect on all causes of death was found for both sexes and age groups; the effects were particularly strong for COPD. The study shows that people with COPD disease and the elderly seem to be affected by air pollution at lower levels than the general population. Important results are available for myocardial infarction as its occurrence [9, 41] and survival have a strong links with particulate matter exposure [42].

This study has some limitations. First, we used proxy measure of exposure at residence rather individual measures of exposure to traffic related pollutants. Second, we used the interpolation method for geocoding instead of a building/parcel method, and this could result in an inaccurate exposure characterization, especially in the periphery of Rome where street segments are longer than in the centre. Finally, when we geocoded the addresses of participants living in a multi-storey building, we assigned the distance to the front door to all residents in the building, irrespective to their real position.

The results of our investigation with respect to the history of previous hospitalization should be interpreted for descriptive purposes only. They are useful to evaluate the specific vulnerability of the population in relation to the environmental exposure and they indicate that from a public health point of view there is a sub-population that is more susceptible that needs to be protected in a more intensive way.

We have shown that living nearby traffic is a widespread phenomenon in Rome, and that people who live close to traffic are older, better educated, and reside in higher socioeconomic areas than those who live far. The study population will be followed to better understand the potential health effects of traffic-related air pollution exposure.