Blood carboxyhæmoglobin levels as a biomarker for urban air pollution exposure

Epidemiological studies have shown that exposure to vehicle air pollution causes a significant threat to human health, especially for vulnerable groups in the population such as young children (Bearer et al., 1995, Gauderman, 2006, Brugge et al., 2007), those undergoing heavy work because of their elevated breathing rate and those suffering from anemia because of their low hemoglobin levels (Hauck and Neuberger, 1984). However, the relationship between air pollution levels (as determined by fixed air pollution monitoring stations) and individual exposure remains poorly understood (di Marco et al., 2005). In addition to the significant limitations associated with determining the actual exposure resulting from people’s movement throughout the day (from home to work or school, etc), there is also limited understanding of the variability in the uptake of pollutants between individuals due to differences in physiology and metabolism. For example, how and to what extent would the uptake of pollutants be different between a group of university students compared to a group of elderly people exposed to the same air pollution levels? And how much inter-person variability can one expect in uptake within specific groups of the population? In this study, we exploit the fact that the uptake of carbon monoxide (a pollutant for which traffic is the dominant source), in the form of blood carboxyhæmoglobin, is reliably described by the established and verified Coburn-Forster-Kane model (Coburn et al., 1965). Much work has been done tovalidate the model in terms of the impact of the various model physiological parameters (such as blood volume, endogenous production of CO, etc) on uptake in the form of sensitivity analyses of the model (Petersen and Stewart, 1970). Despite its age, the CFK model is still regarded as the best all-round model for predicting COHb levels for low-level exposures (WHO, 1999). One of the limitations of the model is that the model physiological variables needed as input to the model (such as the blood volume, endogenous production of CO (produced within the body in the absence of environmental exposure), and the pulmonary diffusing capacity (capacity of the lungs for gas transfer)) are difficult and expensive to measure for large groups and on a routine basis. However, studies have been carried out to relating these model physiological variables and easilymeasured physiological variables (such as weight, height, age and gender) thus avoiding the need for extensive lab-based physiological measurements. For example, it has been established that there is a clear relationship between a person’s weight (and to a lesser extent their age) and their blood volume (Gibson and Evans, 1937). This suggests that if we know the age, gender, weight and height of an individual (or the distribution of these variables within a population) we can estimate their model physiological variables and subsequently make estimates of their expected uptake of CO using the CFK or other suitable model. The first aim of this study is to look at existing results from the literature on the relationship between the CFK model physiological parameters and the so-called ‘easily-measurable physiological parameters’ to determine the reliability in applying the CFK model to epidemiological studies of air pollution within specific populations. The second aim is to evaluate the modelling using a field trial involving human subjects. The results of this study may ultimately help in understanding the health impacts of air pollution exposure of target groups within the population