An examination of cancers diagnosed between 1994 and 2003 showed a higher incidence for all malignant forms of the disease around Dublin and Cork and some other urban areas. Regions with more agricultural workers have a lower risk of disease.

The incidence of all cancers -- excluding non-melanoma skin cancer -- was higher than average in a band running across the north-east and north-midlands from Dublin to Sligo.

The higher incidence of cancer in urban areas is partly linked to social deprivation and lifestyle factors such as smoking and poor diet.

More densely populated areas consistently had a higher risk of cancer than those that were sparsely populated, the report from the National Cancer Registry of Ireland shows.

The risk of almost all cancers analysed was higher in areas with the highest proportion of elderly people living on their own -- this may be due to lifestyle including poorer diet and higher alcohol intake.

The report reveals a higher-than-average incidence of bowel cancer in Cork city and this also extended to the far south-west. It was also higher in the north and centre of the country in a broad band from Dublin through the north-east to Donegal.

The incidence of lung cancer in both sexes was higher in Leinster, most markedly in Dublin, Kildare and Wicklow.

For breast cancer there was a modest variation but it was highest in the major urban areas, with the exception of Limerick. A slight increase was seen in west Cork, north Kerry and large areas in the east-midlands, while within Dublin it was more prominent in the south-east of the capital.

Prostate cancer was highest around the major urban centres, which has been linked to more testing.

Stomach cancer showed the strongest pattern of "clustering" with higher incidence seen in the area stretching from Dublin through Louth, Monaghan and Cavan.

Cervical cancer was seen more around Dublin as well as down the eastern side of the country with a lower incidence in Cork, Kerry and Donegal.

Variations

Commenting on the possible reasons for the variations, the report pointed out that many of the risk factors for cancer were not yet known and therefore the differences could not be fully explained.

The variations do not mean a particular location itself caused the cancer but instead are likely to reflect socio-economic differences in the population, exposure to risk factors and access to or uptake of cancer screening.

"It is generally accepted that it is predominantly the result of differences in well-known risk factors such as tobacco smoking, alcohol drinking, obesity, diet and sexual behaviour" said director Dr Harry Comber.

"We recommend a national research programme to explore, at the individual level, the relationship of known and suspected risk factors to cancer risk and to search for unknown risk factors," he added.