BACKGROUND: In 1991, soil pollution was found around the site of a former chromium-processing factory in Glasgow, Scotland. Levels of chromium in soil were above limits considered as safe, although a risk assessment concluded that population exposure was likely to be below occupational levels. As an excess incidence of leukaemia has been suspected in the area, it was decided to investigate a possible relationship between the pollutant and the illness. METHODS: The ensuing study was descriptive-geographical. In the absence of better data, levels of exposure were assumed to decrease with distance from the centre of the polluted area. Leukaemia and population figures were obtained for each of nine concentric rings by aggregation of data available at the Enumeration District level. The null study hypothesis was that relative risk (as measured by Poisson regression) would not follow a definite trend with distance from the centre. Sex, age and levels of deprivation were taken into account. RESULTS: Relative risks by variables other than distance followed previously known patterns for leukaemia. No evident pattern by distance was found. After regroupings inside the variables, a significant excess of leukaemia was found for intermediate distances from the pollutant. CONCLUSIONS: No evidence was found of a possible relationship between soil pollution by chromium and leukaemia in the general population. Nonetheless, the excess noticed by the study warrants further research.
BACKGROUND: In 1991, soil pollution was found around the site of a former chromium-processing factory in Glasgow, Scotland. Levels of chromium in soil were above limits considered as safe, although a risk assessment concluded that population exposure was likely to be below occupational levels. As an excess incidence of leukaemia has been suspected in the area, it was decided to investigate a possible relationship between the pollutant and the illness. METHODS: The ensuing study was descriptive-geographical. In the absence of better data, levels of exposure were assumed to decrease with distance from the centre of the polluted area. Leukaemia and population figures were obtained for each of nine concentric rings by aggregation of data available at the Enumeration District level. The null study hypothesis was that relative risk (as measured by Poisson regression) would not follow a definite trend with distance from the centre. Sex, age and levels of deprivation were taken into account. RESULTS: Relative risks by variables other than distance followed previously known patterns for leukaemia. No evident pattern by distance was found. After regroupings inside the variables, a significant excess of leukaemia was found for intermediate distances from the pollutant. CONCLUSIONS: No evidence was found of a possible relationship between soil pollution by chromium and leukaemia in the general population. Nonetheless, the excess noticed by the study warrants further research.