The risk of developing many of the cancers presented here was higher in RoI than in NI. The risk of non-melanoma skin cancer, melanoma, leukaemia, bladder, pancreas and brain/central nervous system cancers was significantly higher for both sexes in RoI. For men, the risk of prostate cancer was higher in RoI and, for women, cancer of the oesophagus and cervix. In NI, the risk of lung cancer was higher for both sexes as was that among women for non-Hodgkin’s lymphoma, head and neck cancers and cancer of the uterus. Overall, the relationships between socio-economic variables and cancer risk were similar for men and women. Patterns consistent with known socio-economic gradients were seen—lung, stomach, head and neck and cervical cancers were all more common in areas of higher unemployment and/or lower levels of degree attainment, while non-melanoma skin cancer, female breast cancer, prostate cancer and melanoma were less common. Most cancers were also more frequent in urban areas (as measured by population density); only prostate cancer was more common in rural areas.

Mapping also demonstrated broadly similar geographical patterns for men and women for most cancers. However, apart from this, there was little consistency between different cancer sites in the geographical distribution of risk. There was a marked geographical variation in the risk of some common cancers—non-melanoma skin, lung, prostate and stomach, but very little for others—breast, colorectal, non-Hodgkin’s lymphoma. The most consistent geographical distribution of cancer risk was seen for three cancers (pancreas, brain/central nervous system and leukaemia) which showed an increasing gradient of risk from north-east to south-west.

Eighteen cancer sites were studied, and the results are described in more detail below, in order of cancer frequency.

Non melanoma skin cancer

Non-melanoma skin cancer was 18% commoner in women and 15% commoner in men in RoI than in NI. The risk was higher in more affluent areas, in areas with high levels of elderly living alone, in more densely populated districts and in coastal and urban areas.

Breast cancer

There was no statistically significant difference in female breast cancer risk between RoI and NI. Risk increased with increased population density and affluence. Geographical patterns changed over time, reflecting the introduction of breast screening in RoI in 2000. During 2002-2007, higher rates were seen in the east of RoI (where screening had begun) than in the west (where it had not).

Colorectal cancer

There was no statistically significant difference in colorectal cancer risk between RoI and NI. Increased risk was associated with increasing population density for both sexes and with unemployment for men only. There were areas of higher risk around Cork and from Donegal to Down.

Lung cancer

The risk of lung cancer was significantly higher in NI than RoI for both men (by 11%) and women (by 7%). Increased risk was associated with increased population density, unemployment and low levels of education and was highest in urban areas of Belfast, Dublin, Derry and Cork, and also in Louth, Kildare, Carlow and Wicklow.

Prostate cancer

The risk of diagnosis of prostate cancer was 29% lower in NI than RoI. Men in areas with the highest educational attainment had the highest risk. The risk was highest in the south and east of Ireland during 1995-2001 and in the west and north of RoI during 2002-2007.

Non-Hodgkin’s lymphoma (NHL)

There was no significant difference in risk between NI and RoI for men but the risk for women was 14% higher in NI. There was no association between NHL risk and population density or socio-economic factors. The highest risk for men was in the north-east, and in Kerry and Galway and in the north-east and Dublin for women.

Stomach cancer

Stomach cancer risk was higher in areas of high population density and in those with high unemployment and lower educational attainment. There was a strong geographical pattern, with higher risk in a band running from Dublin to Donegal, excluding the north-east, but including Belfast.

Melanoma of the skin

Melanoma risk was lower in NI than RoI for both men (by 8%) and women (by 14%). Risk was not associated with population density; but was associated, for both men and women, with low unemployment and high educational attainment, and was highest in coastal areas in the south and east of Ireland.

Bladder cancer

Bladder cancer risk was lower in NI for men (by 8%) and for women (by 14%) than in RoI and increased with population density (for both sexes) and unemployment (men only). Male geographical patterns were distinctive, with increased risks from Louth to Wicklow, including Dublin city, and also in Donegal, North Down and Ards. For women, there was an area of higher risk in the south-west.

Head and neck cancer

The risk of head and neck cancer was greater, by 21%, for women in NI compared to RoI but there was no statistically significant difference for men. The risk increased with increased population density and unemployment but not with educational attainment. There was no clear geographical pattern for men, but for women there was one large area of higher relative risk stretching north-westwards in a line between Dublin and Sligo.

Leukaemia

The risk of leukaemia was lower in NI than in RoI, by 23% for men and 17% for women. There was no association with population density, employment or educational attainment. Mapping showed an increasing gradient of increasing risk from north-east to south-west, more pronounced in men.

Pancreatic cancer

Pancreatic cancer risk was lower in NI than in RoI, by 11% in men and 22% in women. Increased risk was associated with higher unemployment, but only for men. In women the risk increased with decreasing levels of educational attainment. There was a gradient across the island, with increased risk in the south-west and lowest in north-east; this pattern was more marked in women than men.

Kidney cancer

There was no statistically significant difference in risk between NI and RoI and no association with either population density or socio-economic factors. The area of highest risk was mainly in Leinster, with a lower relative risk in the west.

Oesophageal cancer

The risk was 8% lower in NI than RoI for women but there was no statistically significant difference for men. Risk increased with increasing population density for both sexes but there was no association with unemployment or educational attainment. The area of highest risk was south of a line from Dublin to Kerry, with areas of low risk in the north-west.

Ovarian cancer

There was no association with country, population density or socio-economic factors. The areas of highest risk were around Cork city, extending more widely across most of Munster and also in the eastern half of NI, excluding parts of Down and Belfast.

Brain and other central nervous system (CNS) cancers

The risk of brain and other CNS cancer was lower in NI than RoI, by 10% for men and 20% for women. There was a weak positive association with population density for women only, and no association with socio-economic variables for either sex. There was a strong geographical pattern, with the highest risk in the south-west and lowest risk in the north-east.

Cancer of the corpus uteri

The risk of cancer of corpus uteri was higher by 11% in NI than in RoI. There was no significant association with population density or socio-economic variables. Higher risk was noted in Connacht, most of Northern Ireland and parts of Munster.

Cancer of the cervix uteri

The risk of cancer of the cervix was significantly higher, by 11%, in RoI than NI and increased with increasing population density, unemployment and poor educational attainment. The areas of highest risk were around Dublin, extending to Wexford and the midlands. Areas around Cork, Waterford, Belfast and Sligo also had higher risk.

This atlas shows major variations, sometimes more than two-fold, in the risk of several cancers, across the island. For many cancers, we found a strong relationship between markers of socio-economic status and cancer risk, sometimes positive, sometimes negative. These socio-economic relationships were more consistent than the broad geographical patterns identified by mapping. Few of the geographical patterns could be satisfactorily explained by the available data on risk factors, although we did see some correlations between smoking prevalence and smoking-related cancers.

Some differences in relative risk appeared to be attributable to health service provision—higher levels of breast screening in NI in the 1990s; more prostate specific antigen testing in RoI. For a few cancers, the more demand-led service in RoI may lead to more case-finding and an apparently higher overall cancer risk, as seen in its most extreme form for prostate cancer.

We were struck by the relative paucity of comparable information on established cancer risk factors at individual, small area or national level in both countries. Sources of data are fragmented and often either unavailable or not published. Understanding the reasons for geographical variation, and taking appropriate action, would reduce the cancer burden significantly in Ireland. We hope that this atlas will serve as a stimulus and raw material for detailed studies which will explore and answer some of the questions it poses.

Research

1. Areas with unexplained higher than average risk should be studied in detail. Some findings here are of major public health importance:
   • The largely unexplained geographical patterns for some of the more common cancers: colorectal, melanoma and stomach cancer;
   • The role of H pylori infection, an easily diagnosed and treatable risk factor, in the high rates of stomach cancer in the east and north-west.
2. A comprehensive programme of research into already known determinants of cancer risk in Ireland is needed to inform cancer control. Information on modifiable risk factors such as smoking, diet, exercise, alcohol use, medication use, reproductive history and infection, their population prevalence and variation by age, sex, socio-economic status, area of residence, and over time, is essential for an understanding of the cancer burden.
3. The aetiology of cancers where risk factors are uncertain requires further investigation, ideally through international collaborations.
4. Patterns of healthcare access and utilisation in Ireland, north and south, and how these affect cancer risk, need to be understood.

Risk reduction

1. Research is needed into levels of awareness and knowledge of cancer risk factors among the population in Ireland, and how these vary by age, sex, socio-economic status and geographical area.
2. Action should be taken to reduce the prevalence of important modifiable risk factors such as tobacco, alcohol, overweight and UV exposure.
3. Campaigns and initiatives to help raise awareness of “healthy” lifestyle behaviours (e.g. physical activity) among the public should emphasise the links between lifestyle and cancer.

Data and linkage

1. Collection of information on determinants of health (including socio-economic data collected as part of the census) in the population of both countries should be consistent, integrated, systematic and regular.
2. Efforts should be made to increase the compatibility of health, risk factor, census and health service utilisation data between NI and RoI.
3. Both countries should run regular, comparable, population-based health surveys and make the data available for research. Linkage of this data with other routinely collected data should not be unnecessarily restricted.
4. Public health data which is collected with the support of public funding should be made readily available to researchers, with appropriate safeguards.
5. Custodians of data relevant to health should be encouraged to clarify the potential for data linkage and the use of this data for the public good should be encouraged by Government policy.

From the Republic of Ireland perspective, the island of Ireland, and the country comprising most of its area, are both officially known as “Ireland”, while in Northern Ireland there is no universally agreed terminology for the different geographic bodies in the island of Ireland. As this is certain to cause confusion in an atlas of this kind, we have elected to use the expressions “Republic of Ireland (RoI)” and “Northern Ireland (NI)” for the two jurisdictions on the island and to refer to both of these areas as “countries”. The combined area of the whole island (and the offshore islands) is referred to in the text simply as “Ireland”. None of this implies anything concerning the official status of these names.

For administrative purposes, RoI is divided into 27 counties (including Tipperary North and South) (Map 2.1). Counties with large urban areas are further divided into “city” and “county” areas (three of the latter in the case of Dublin), giving a total of 34 large administrative areas. Small area population statistics are available at the level of electoral division (ED), of which there are approximately 3,500.

NI has six counties and is also divided into 26 district councils—four city councils (Armagh, Belfast, Derry and Lisburn), 13 Borough Councils and 9 District Councils (Map 2.1). Small area population statistics are available at the level of ward, of which there are approximately 580. District councils in NI therefore have smaller populations on average than counties in RoI, while NI wards have an average population size greater than RoI electoral divisions (see section 2.2.4). Mapping and statistical analysis in this atlas is based almost exclusively on data at the ward/ED level.

A number of other geographical entities are also referred to in the atlas—these include health board areas, health service regions, provinces and planning regions. These are described in Appendix table A4.1.

Map 2.2 shows, for reference, the outlines of counties in RoI and District Councils in NI with some towns and cities. 

Map 2.1 Counties and district councils

Map 2.2 Locations

Non-melanoma skin cancer (NMSC) was the most common cancer in Ireland, accounting for 27% of all malignant neoplasms (Table 3.1). The average number of new cases diagnosed each year was 3,777 in women and 4,294 in men. During 1995-2007, the number of new cases increased by approximately 3% per annum; since 2002 it has been increasing by around 6% in RoI.

The risk of developing NMSC before the age of 75 was 1 in 12 for women and 1 in 8 for men and was slightly higher in RoI than in NI for both men and women. At the end of 2008, 11,629 women and 12,375 men aged under 65, and 30,748 women and 31,937 men aged 65 and over, were alive up to 15 years after their cancer diagnosis.

Table 3.1 Summary information for non-melanoma skin cancer in Ireland, 1995-2007

The incidence of NMSC increased with increasing age (Figure 3.1). The age distribution was similar for men and women and for RoI and NI. Only approximately 10% of cases occurred in those aged under 50 years and the largest number of cases for both sexes presented in the 70–79 age group.

Figure 3.1 Age distribution of non-melanoma skin cancer cases in Ireland, 1995-2007, by sex

No reliable data are available on international variations in non-melanoma skin cancer incidence.

Table 3.2 Risk factors for non-melanoma skin cancer, by direction of association and strength of evidence

Individuals who are immune suppressed, such as organ transplant recipients or those with AIDS, have a greatly increased risk of developing NMSC. Positivity for the human immunodeficiency virus, type 1 (HIV-1) is a cause of NMSC. Some immunosuppressive drugs—including azathioprine and ciclosporin—which are used to prevent organ rejection following transplant, or to treat autoimmune diseases such as rheumatoid arthritis and Crohn’s disease, are recognised to cause skin cancer. Risk of NMSC is also increased by exposure to the drug methoxsalen, which is used to treat some skin conditions, in combination with UV light. Residues of arsenic from agriculture, mining and industrial practices can end up in drinking water. Arsenic is carcinogenic (International Agency for Research on Cancer, 1987; International Agency for Research on Cancer, 2004a) and ingestion of arsenic and inorganic arsenic compounds causes NMSC. Low-dose ionizing radiation exposure (e.g. for benign skin conditions such as acne) increases risk of BCC.The two main types of non-melanoma skin cancer are squamous cell carcinoma (SCC) and basal cell carcinomas (BCC). Both types are caused by exposure to ultraviolet (UV) radiation present in sunlight. Occupational sunlight exposure has been mainly associated with SCC and recreational exposure with BCC. Individuals with a lighter skin colour, less ability to tan, and who had freckles as a child, are at increased risk, as are those with solar keratoses (benign sun damage to the skin). Independently of sun exposure, use of artificial tanning devices which emit UV radiation, such as sunbeds or sunlamps, has been associated with raised risk of BCC and, especially, SCC.

Human papilloma viruses (HPV) infect mucosal and cutaneous epithelia. Infection with particular HPV types (genus-beta types and specifically HPV5 and HPV8) may be causally related to NMSC (International Agency for Research on Cancer, 2011b). People who use statins may have an increased risk of NMSC (although the possibility that the association could be due to different levels of contact with health services among users and non-users of statins cannot be discounted).

Non-melanoma skin cancer had a strong geographical pattern, which was similar for men and women (Maps 3.1-3.3).

Regions of high relative risk were mainly seen in coastal areas, particularly along the east coast from Down to Wicklow, the south and west coasts from Waterford to Mayo and in Sligo (men) and Donegal. Areas of higher relative risk were also seen around the cities of Dublin, Waterford, Cork, Limerick and Galway. 

Map 3.1 Non-melanoma skin cancer, smoothed relative risks: both sexes

Map 3.2 Non melanoma skin cancer, smoothed relative risks: males

Map 3.3 Non melanoma skin cancer, smoothed relative risks: females

The average number of new cases diagnosed each year was 2,965 in women and 20 in men. Breast cancer was the most common cancer in women in Ireland, accounting for 29% of all malignant neoplasms, excluding non-melanoma skin cancer (Table 4.1).  During 1995-2007, the number of new cases diagnosed in women increased by approximately 3% per annum, 4% in RoI and 1% in NI. From 2002 to 2007, annual increases of 3% in RoI and 4% in NI were observed.

The risk of developing breast cancer up to the age of 74 was 1 in 12 for women and 1 in 1,621 for men and was slightly higher in RoI than in NI. At the end of 2008, 17,167 women and 53 men aged under 65, and 13,987 women and 128 men aged 65 and over, were alive up to 15 years after their cancer diagnosis.

Table 4.1 Summary information for breast cancer in Ireland, 1995-2007

The remainder of this chapter relates only to breast cancer in women.

The proportion of breast cancers which occurred in women aged 60 and over was lower than for most other cancers (Figure 4.1). Almost one quarter of all cases occurred in women aged under 50, and a further quarter in those aged 50–59. Just 12% of cases were diagnosed in those aged over 80 years. This pattern was similar for RoI and NI.

Figure 4.1 Age distribution of female breast cancer cases in Ireland, 1995-2007

The incidence of breast cancer in women in RoI and NI was close to the median of the 21 countries shown. Age-standardised rates in both countries were slightly lower than in the UK (Figure 4.2). Rates were highest in Belgium and Denmark and lowest in Poland, Russia and Japan.

Figure 4.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: female breast cancer

Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from cancer registry data for 2005-2007)

Table 4.2 Risk factors for breast cancer, by direction of association and strength of evidence

Breast cancer is a heterogeneous disease, comprising several distinct subgroups defined on the basis of hormonal receptor status and/or morphology. Recently interest has grown in distinguishing between risk factors for different subtypes (see, for example, Suzuki et al., 2008; Reeves et al., 2009; Suzuki et al., 2009; Vrieling et al., 2010; Yang et al., 2011). Up to 10% of breast cancer cases are hereditary and a woman's chance of developing the disease is increased if any of her first degree female relatives had breast cancer, particularly if more than one relative was affected at a young age (Veronesi et al., 2005). By age 70, women who carry BRCA1 gene mutations have a 65% chance of developing breast cancer, while those who carry BRCA2 mutations have a 45% risk (Antoniou et al., 2003). Family history may interact with other factors to modify risk, for example, exposure to low doses of radiation such as x-rays (Jansen-van der Weide et al., 2010) or history of benign breast disease (Zhou et al., 2011). Other than genetic factors, the major determinant of breast cancer risk is lifetime exposure to oestrogen (Table 4.2). Higher endogenous oestrogen exposure, as well as exogenous oestrogens, increases risk. In contrast, in pre-menopausal women at high risk of breast cancer, the anti-oestrogenic drugs tamoxifen and raloxifene reduce the chances of developing the disease by about half. 

Colorectal cancer was the second most common cancer in Ireland. It accounted for 13% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 15% in men (Table 5.1). The annual average number of new cases diagnosed was 1,307 in women and 1,631 in men. 63% of these cancers arose in the colon and 37% in the rectum. During 1995-2007, the number of new cases increased, on average, by approximately 2% per annum, but during 2002-2007 increased by 4% per annum. The patterns of increase were similar for RoI and NI.

The risk of developing colorectal cancer up to the age of 74 was 1 in 32 for women and 1 in 20 for men and was similar for NI and RoI. At the end of 2008, 2,335 women and 2,787 men aged under 65, and 6,137 women and 7,205 men aged 65 and over, were alive up to 15 years after their colorectal cancer diagnosis.

Table 5.1 Summary information for colorectal cancer in Ireland, 1995-2007

Approximately half of colorectal cancers were diagnosed in those aged 70 and older—51% of men and 58% of women (Figure 5.1). The age distribution was similar for both sexes, although there was a higher proportion of cases in men aged 60–69 (29%, compared to 22% in women) and a higher proportion in women aged 80 and older (26%, compared to 17% in men).

Figure 5.1 Age distribution of colorectal cancer cases in Ireland, 1995-2007, by sex

The incidence of colorectal cancer among both men and women in RoI and NI was close to the median for the countries shown. Rates were highest in New Zealand among women and in the Czech Republic for men (Figure 5.2).

Figure 5.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: colorectal cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from cancer registry data for 2005-2007)

Table 5.2 Risk factors for colorectal cancer, by direction of association and strength of evidence

Up to 10% of colorectal cancers are hereditary and most are due to the genetic syndromes of familial adenomatous polyposis (FAP) and hereditary non-polyposis colorectal cancer (HNPCC) (Hawkins and Ward, 2001). Excluding these syndromes, individuals who have a first degree relative with colorectal cancer have around a two-fold increased risk of developing the disease themselves.

Lifestyle factors are extremely important in colorectal cancer (Table 5.2). Smoking is causally related to colon, but not rectal, cancer. Alcohol is a cause of both colon and rectal cancers. Higher levels of body fatness, and in particular central adiposity, are positively related to risk. In a recent meta-analysis, each 5kg/m² increment in body mass index was associated with an 18% increase in risk; the association appears stronger for colon than rectal cancer, for men than women, and in studies adjusting for physical activity (Ning et al., 2010). In contrast, physical activity is consistently inversely associated with colon cancer, in particular, and risk decreases in a dose-response fashion with increased frequency or intensity of activity. Regular use of aspirin or other non-steroidal anti-inflammatory drugs may reduce colorectal cancer risk by up to half. In addition, risk is decreased in women taking hormone replacement therapy and is likely also to be lower in those who have taken oral contraceptives.

Many studies have found increased risk in individuals who have higher intakes of processed meats (preserved by smoking, curing or salting, such as ham, bacon or salami) and red meats. In contrast, higher intake of various other dietary components may be associated with lower risk, including garlic; fruit; fish; non-starchy vegetables; milk, dairy products or calcium; coffee; soya and soya foods; and foods containing dietary fibre or the B vitamin folate.

Map 5.1 Colorectal cancer, smoothed relative risks: both sexes

Map 5.2 Colorectal cancer, smoothed relative risks: males

Map 5.3 Colorectal cancer, smoothed relative risks: females

Lung cancer was the third most common cancer in Ireland, accounting for 10% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 15% in men (Table 6.1). The average number of new lung cancer cases diagnosed each year was 1,000 in women and 1,602 in men. During 1995-2007, the number of new cases diagnosed increased by approximately 3% per annum for women and 1% for men.

The risk of developing lung cancer up to the age of 74 was 1 in 37 for women and 1 in 20 for men, and was higher in NI than in RoI for both males and women. At the end of 2008, 708 women and 768 men aged under 65, and 1,330 women and 1,577 men aged 65 and over, were alive up to 15 years after their lung cancer diagnosis.

Table 6.1 Summary information for lung cancer in Ireland, 1995-2007

Lung cancer is mostly a disease of older people—the majority of cases occurred in those aged 70 and over (Figure 6.1). Less than 5% of cases presented in persons aged under 50.

Figure 6.1 Age distribution of lung cancer cases in Ireland, 1995-2007, by sex

Incidence rates for lung cancer varied considerably among developed countries, particularly among men (Figure 6.2). Poland demonstrated the highest male rates, while USA, Denmark and Canada had the highest female rates. Rates were lowest among men in Sweden, and in Portugal and Russia among women. Lung cancer risk in both RoI and NI was close to the median compared to other developed countries for men, but was slightly above the median for women.

Figure 6.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: lung cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from cancer registry data for 2005-2007)

Table 6.2 Risk factors for lung cancer, by direction of association and strength of evidence

Smoking is the principal cause of lung cancer (Table 6.2). In populations with prolonged cigarette use, 90% of lung cancer cases are due to cigarette smoking (International Agency for Research on Cancer, 2004b). Risk increases with younger age at smoking commencement and longer duration of smoking. Stopping smoking, at any age but particularly so before middle age, avoids most of the subsequent risk (Peto et al., 2000). In addition, involuntary exposure to tobacco smoke (passive smoking) is a cause of lung cancer in those who have never smoked. The consistent relationship between higher lung cancer risk and lower socio-economic status probably reflects social class variations in tobacco exposure.

The chances of developing lung cancer are increased in those exposed to radon, ionizing radiation and a variety of chemicals and compounds. Various lifestyle factors (such as alcohol intake, physical activity, leanness and aspects of diet) may be related to lung cancer, but it is not always possible to rule out the possibility that the associations are due to a residual effect of smoking. 

Map 6.1 Lung cancer, smoothed relative risks: both sexes

Map 6.2 Lung cancer, smoothed relative risks: males

Map 6.3 Lung cancer smoothed relative risks: females

Prostate cancer was the most common cancer in men, accounting for 23% of all malignant neoplasms, excluding non-melanoma skin cancer, in men in Ireland (Table 7.1). The average number of new cases diagnosed each year was 2,550. Between 1995 and 2007, the number of new cases more than doubled, increasing by approximately 7% per annum. From 1995 to 2001, the number of new cases diagnosed per annum rose by approximately 9% in RoI and by 2% in NI. From 2002 to 2007 the annual increase was 5% in RoI and 6% in NI.

The risk of developing prostate cancer up to the age of 74 was 1 in 13 and was higher in RoI than in NI. At the end of 2008, 5,235 males aged under 65 and 17,829 aged 65 and over were alive up to 15 years after their prostate cancer diagnosis.

Table 7.1 Summary information for prostate cancer in Ireland, 1995-2007

Prostate cancer is predominantly a disease of old age. Just over 1% of cases presented in those aged under 50, while 87% occurred in those aged 60 and older (Figure 7.1). Age at diagnosis was slightly older in NI than in RoI, with 60% aged 70 and over at diagnosis in NI compared with 53% in RoI.

Figure 7.1 Age distribution of prostate cancer cases in Ireland, 1995-2007

Considerable variation exists in prostate cancer incidence rates among developed countries. Estimates for 2008 ranged from 105 per 100,000 men in Australia to 23 per 100,000 in Japan (Figure 7.2). The rate in RoI was in the upper third of the range, while that in NI was in the lower third.

Figure 7.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: prostate cancer

Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from cancer registry data for 2005-2007)

Table 7.2 Risk factors for prostate cancer, by direction of association and strength of evidence

The chances of a man developing prostate cancer are approximately doubled if he has a relative with the disease. Despite extensive study, relatively few other risk factors for prostate cancer have been firmly established (Table 7.2). Taller men appear to have higher risk, although height per se is unlikely to affect a man’s chances of developing the cancer: instead it probably points to relevant childhood exposures. One possibility is that the association operates through the insulin-like growth factor system, which plays a key role in cell proliferation, differentiation and apoptosis. Positive associations have been reported between prostate cancer and circulating levels of insulin-like growth factor 1.

There is limited evidence that exposure to arsenic and inorganic arsenic compounds (through occupational inhalation or ingestion in food or drinking water) and cadmium and cadmium compounds (via occupational use), are a cause of prostate cancer.

Some aspects of diet may influence risk. Higher intake of the carotenoid lycopene is associated with decreased risk in many studies. Risk may be reduced in men, especially those from Asian populations, who consume more tofu and soya food products. While observational studies suggest that higher intake of selenium or selenium-rich foods is also inversely associated with prostate cancer, two trials found no effect of selenium supplementation.

Obesity may be associated with reduced risk of non-aggressive prostate cancer but, since the opposite has been suggested for aggressive disease, the observations could be due to a detection bias relating to the ability to detect prostate cancer in obese men (Buschemeyer and Freedland, 2007). While current smoking per se does not appear to be associated with risk, risk may be elevated among smokers with higher levels of exposure. Meta-analyses suggest the possibility that regular use of aspirin and other non-steroidal anti-inflammatory drugs could be associated with reduced risk of prostate cancer overall, and of advanced cancers specifically, but there are inconsistencies and limitations in the evidence. 

Map 7.1 Prostate cancer, smoothed relative risks: 1995-2007

Map 7.2 Prostate cancer, smoothed relative risks: 1995-2001

Map 7.3 Prostate cancer, smoothed relative risks: 2002-2007

Non-Hodgkin’s lymphoma (NHL) consists of a group of more than 20 different malignant lymphoproliferative diseases that originate from lymphocytes. NHL was the fifth most common cancer in Ireland, accounting for 3.4% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 3.6% in men (Table 8.1). The average number of new cases diagnosed each year was approximately 354 in women and 392 in men. During 1995-2007, the number of new cases diagnosed increased by approximately 3% per annum overall.

The risk of developing NHL up to the age of 74 was 1 in 102 for women and 1 in 80 for men, and was higher in NI than in RoI for both males and females. At the end of 2008, 1,255 females and 1,607 males aged under 65, and 1,477 females and 1,266 males aged 65 and over, were alive up to 15 years after their NHL diagnosis.

Table 8.1 Summary information for non-Hodgkin’s lymphoma in Ireland, 1995-2007

Age at diagnosis of NHL was younger for men than for women—43% of cases presented in men under 60 compared to 33% of women (Figure 8.1). Approximately 17% of women and 10% of men were aged 80 years or older at diagnosis. Age at diagnosis was also slightly younger in RoI than in NI.

Figure 8.1 Age distribution of non-Hodgkin’s lymphoma cases in Ireland, 1995-2007, by sex

Among developed countries the USA had the highest incidence rate of NHL for men and women, while Canada had the second highest rates for both sexes. Both RoI and NI had close to median rates of the disease for men and women, while Poland, Russia and Japan had the lowest rates.

Figure 8.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: non-Hodgkin’s lymphoma
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007) NOTE: NHL was defined in GLOBOCAN by ICD10 codes C82-C85 & C96. In the atlas we have included codes C82-C85 only in the definition of NHL (see Table 2.1.) The data for NI and RoI in Figure 8.2 retain this definition, while those for other countries include C96. Due to changes in case definition, inclusion of C96 in the RoI data would have artificially inflated the incidence figures.

Table 8.2 Risk factors for non-Hodgkin’s lymphoma, by direction of association and strength of evidence

Although lymphomas may occur in children, the following description of risk factors relates to NHL in adults.

The best described risk factor for NHL is immune deficiency (Table 8.2). Individuals who have had organ transplants, or who take immunosuppressive drugs to prevent organ rejection or to treat various autoimmune conditions (e.g. azathioprine and ciclosporin) have a high risk of NHL. Risk is also increased in individuals with some auto-immune disorders, including Sjorgen syndrome, systemic lupus and haemolytic anaemia. A variety of infection—including human immunodeficiency virus, type 1 (HIV-1) which causes AIDS, Epstein-Barr virus (EBV), which is a common member of the herpesvirus family, and hepatitis C virus (HCV)—can cause NHL, often acting via effects on immunosuppression or inflammation.

Occupational exposure to ethylene oxide (an organic compound used in the manufacture of products like polyethylene), benzene (an industrial solvent and precursor to basic industrial chemicals including drugs, plastics, synthetic rubber, and dyes) and formaldehyde (used in the production of industrial resins that are then used in the manufacture of products such as adhesives and binders for wood products, plastic and synthetic fibres) are all associated with NHL. Those exposed to pesticides through their work, or to tricholoroethylene, a hydrocarbon commonly used as an industrial solvent, may also have increased risk of NHL, but the evidence is less certain. Workers in the rubber manufacturing industry are also at increased risk, but due to the complexity of exposures in this industry, the causative agents have not yet been identified.

Risk of NHL is increased by around 50% in those who report first-degree relatives with NHL, Hodgkin’s lymphoma or leukaemia. In terms of lifestyle factors, current smoking may increase the risk of follicular lymphomas, but not other subtypes. A modest positive association between weight and NHL risk is evident in meta-analyses, but this may be limited to particular subtypes and/or to obese individuals. Increased recreational sun exposure may be associated with a modestly reduced risk of particular types of NHL, while individuals with allergic conditions, such as hay fever or asthma, or atopy may have slightly reduced risk of B-cell NHL. 

Map 8.1 Non-Hodgkin’s lymphoma, smoothed relative risks: both sexes

Map 8.2 Non-Hodgkin’s lymphoma, smoothed relative risks: males

Map 8.3 Non-Hodgkin’s lymphoma, smoothed relative risks: females

Stomach cancer was the sixth most common cancer in Ireland, accounting for 2.7% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 4.0% in men (Table 9.1). The average number of new cases diagnosed each year was 278 in women and 442 in men. During 1995-2007, the number of new cases diagnosed per annum remained fairly constant.

The risk of developing stomach cancer up to the age of 74 was 1 in 161 for women and 1 in 74 for men and was similar in NI and RoI for both men and women. At the end of 2008, 197 women and 323 men aged under 65, and 522 women and 773 men aged 65 and over, were alive up to 15 years after their stomach cancer diagnosis.

Table 9.1 Summary information for stomach cancer in Ireland, 1995-2007

More than 50% of all cases of stomach cancer were diagnosed at over 70 years of age—65% of women and 53% of men (Figure 9.1). Only 7% of cases were aged under 50 years at diagnosis. Patterns were similar for RoI and NI, but with a slightly older age at diagnosis in NI.

Figure 9.1 Age distribution of stomach cancer cases in Ireland, 1995-2007, by sex

Rates of stomach cancer among men and women were considerably higher in Japan than in other developed countries. Within Europe, Russia had the highest incidence rate (Figure 9.2). In RoI and NI rates were moderate compared to other countries for both men and women. Sweden, USA and Canada had particularly low rates of this cancer.

Figure 9.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: stomach cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007).

Table 9.2 Risk factors for stomach cancer, by direction of association and strength of evidence

Infection with the common bacterium, Helicobacter pylori (H pylori), which lives in the stomach and causes inflammation and ulcers, is associated with a six-fold raised risk of stomach cancer (Table 9.2). Meta-analysis of intervention studies shows that stomach cancer risk is decreased by one-third in H pylori-positive patients randomised to eradication treatment (Fuccio et al., 2009). It has been suggested that H pylori infection may be a necessary (but not sufficient) cause of cancers in the distal stomach (International Agency for Research on Cancer, 1994). It is also probable that infection with the Epstein-Barr virus (EBV), which is a member of the herpesvirus family, and very common, is a cause of stomach cancer.

Smoking is firmly established as a cause of stomach cancer. Compared to non-smokers, risk in increased by 50% in those who have ever smoked and 70% in current smokers (La Torre et al., 2009). Risk also increases with number of cigarettes smoked (Ladeiras-Lopes et al., 2008). Those with low socio-economic status have increased stomach cancer risk, probably in part reflecting variations in tobacco use by social class.

Other than these factors, the main risk factors are related to food and food preservation. There is substantial and consistent evidence that higher intakes of salt, salty foods or foods preserved in salt are associated with increased risk. Risk is reduced in individuals with higher intakes of fruit and non-starchy or fresh vegetables; in contrast there may be a modest increased risk in those who consume higher quantities of pickled vegetables. More than 10 studies have reported a significant reduction in disease risk with use of refrigeration. However, it is thought that the association is not due to refrigeration per se but rather is a consequence of other factors related to refrigerator use, such as lower intake of foods preserved with salt, or higher intake of fresh perishable foods (e.g. vegetables and fruit) (World Cancer Research Fund / American Institute for Cancer Research, 2007).

In terms of other potential risk factors, use of aspirin and other non-steroidal anti-inflammatory drugs has been associated with a modest (approximately 20%) reduction in risk of stomach cancer; this association appears stronger after adjusting for other risk factors. Overweight individuals (body mass index >25kg/m²) may have a modest increased risk, particularly for cancers arising in the gastric cardia.

Stomach cancer had a strong geographical pattern with lower relative risk in most of the south and west of Ireland and higher relative risk in parts of the north (Maps 9.1-9.3).

For both sexes combined, the main area of higher relative risk extended northwards from Dublin as far as the border counties and west into the midlands (Map 9.1). There were also isolated areas of higher risk in Donegal, Belfast, Antrim, north Lisburn, Kerry and Connacht.

The pattern for men was similar to that for both sexes combined (Map 9.2).

For women, the main area of higher relative risk extended in a band from Dublin to Donegal, excluding the north-east of NI, but including Belfast, south Antrim, north Lisburn, Ards and parts of Down. Comparatively high relative risk was also found in western parts of Galway, west Cork and west Kerry (Map 9.3).

Map 9.1 Stomach cancer, smoothed relative risks: both sexes

Map 9.2 Stomach cancer, smoothed relative risks: males

Map 9.3 Stomach cancer, smoothed relative risks: females

Melanoma of the skin was the seventh most common cancer in Ireland, accounting for 4.1% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 2.6% in men (Table 10.1). The average number of new cases diagnosed each year was 421 in women and 285 in men. During 1995-2007, the number of new cases diagnosed increased at approximately 5% per annum overall.

The risk of developing melanoma up to the age of 74 was 1 in 89 for women and 1 in 116 for men, and was slightly higher in RoI than in NI. At the end of 2008, 2,991 women and 1,707 men aged under 65, and 2,140 women and 1,336 men aged 65 and over, were alive up to 15 years after their melanoma diagnosis.

Table 10.1 Summary information for melanoma of the skin in Ireland, 1995-2007

Melanoma of the skin is a disease of younger age groups, with 36% of women and 31% of men aged under 50 at diagnosis (Figure 10.1). Over half of all new cases presented under 60 years. However, a substantial proportion of cases (12-13%) was diagnosed in the age group 80 and older. Age patterns were similar for RoI and NI.

Figure 10.1 Age distribution of cases of melanoma of the skin in Ireland, 1995-2007, by sex

For both men and women Australia and New Zealand had the highest incidence rates of melanoma, while Japan had the lowest (Figure 10.2). Both RoI and NI had moderate rates of melanoma compared to other developed countries. Rates in both parts of Ireland were lower than in the UK.

Figure 10.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: malignant melanoma of skin
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007).

Table 10.2 Risk factors for melanoma of the skin, by direction of association and strength of evidence

Exposure to ultraviolet (UV) radiation, the primary source of which is sunlight, is the main cause of melanoma of the skin (Table 10.2). Risk increases with increasing levels of intermittent, or recreational, sun exposure. A history of sunburn, often considered to be a marker of high levels of intermittent sun exposure, is associated with raised risk in a dose-response fashion. Presence of benign sun damage (solar keratoses) is a marker of increased risk, particularly for melanomas on the head and neck. Exposure to artificial UV radiation, through use of sunbeds or sunlamps, also causes melanoma. Constitutional factors act together with UV exposure to influence the chance of an individual developing melanoma. Risk is increased in those with more naevi (moles), a high density of freckles (or who had freckling as a child), light hair, skin or eye colour, and less ability to tan.

Melanoma risk is higher in those of higher socio-economic status. This may be due to greater recreational sun exposure among more affluent groups.

Individuals with one or more first degree relatives who have had melanoma have a two-fold increased risk of developing it themselves. On average, around 4% of cases of melanoma are estimated to be attributable to familial risk.

In terms of more tentative associations, modest positive relationships have been reported between melanoma risk and higher body mass index in men and weight gain in adulthood in women. 

Melanoma of the skin had a strong geographical pattern which was similar for men and women. Areas of higher relative risk were seen along the south coast from Cork to Wexford, and also along the east coast from Meath to Wicklow and in parts of Larne, Castlereagh, Ards, Down, Armagh, Craigavon, Banbridge, Newry and Mourne, Kerry, Limerick, Galway and Mayo (Map 10.1).

The pattern for men was similar to that for both sexes combined (Map 10.2)

For women (Map 10.3) the area of higher risk was slightly more dispersed around Dublin. Areas of higher relative risk were also seen in western parts of Galway and Mayo and in north Leinster.

Map 10.1 Melanoma of the skin, smoothed relative risks: both sexes

Map 10.2 Melanoma of the skin, smoothed relative risks: males

Map 10.3 Melanoma of the skin, smoothed relative risks: females

Bladder cancer was the fourth most common cancer in Ireland for men and the twelfth most common for women, accounting for 1.9% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 4.4% in men (Table 11.1). The average number of new cases diagnosed each year was 193 in women and 479 in men. During 1995-2007, the number of new cases diagnosed per annum remained fairly constant.

The risk of developing bladder cancer up to the age of 74 was 1 in 212 for women and 1 in 72 for men and was slightly higher in RoI than in NI. At the end of 2008, 326 women and 787 men aged under 65, and 981 women and 2,527 men aged 65 and over, were alive up to 15 years after their bladder cancer diagnosis.

Table 11.1 Summary information for bladder cancer in Ireland, 1995-2007

The distribution of age at diagnosis for bladder cancer was similar for men and women. Almost 60% of new cases were aged 70 or older at diagnosis (Figure 11.1). Only about 6% of cases presented at under 50 years of age. Age at diagnosis was slightly younger in RoI than in NI.

Figure 11.1 Age distribution of bladder cancer cases in Ireland, 1995-2007, by sex

Male and female rates of bladder cancer in NI were among the lowest of the 21 developed countries examined, with RoI also having a comparatively low incidence in men but not women (Figure 11.2). Female rates were much lower overall than male rates, and varied somewhat less. Spain had the highest rate of bladder cancer among men, while Denmark had the highest rate among women. However, caution needs to be exercised in making international comparisons of bladder cancer incidence, as classification of the disease with regard to malignant and non-malignant cancers is not consistent across countries.

Figure 11.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: bladder cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 11.2 Risk factors for bladder cancer, by direction of association and strength of evidence

The leading known cause of cancer of the bladder is tobacco smoking (Table 11.2). Two-thirds of bladder cancers in men and one-third in women are considered to be due to smoking (Brennan et al., 2000; Brennan et al., 2001). Risk increases with duration of cigarette smoking and number of cigarettes smoked (International Agency for Research on Cancer, 2004b). Risk is also increased in those who smoke pipes or cigars, but do not smoke cigarettes (Pitard et al., 2001). Stopping smoking results in an immediate decrease in risk (Scélo and Brennan, 2007).

A range of occupations (including painting, mining, bus driving, motor mechanic, blacksmith, machine setter and hairdressing) and employment in various industries or in manufacturing of specific products (including rubber manufacturing, aluminium production and magenta manufacture) are associated with bladder cancer. In terms of specific occupational exposures which cause bladder cancer, the most consistent evidence relates to aromatic amines and polycyclic aromatic hydrocarbons.

Other than smoking and occupational exposures, the factors involved in bladder cancer aetiology are largely unknown, although several putative relationships have been suggested. Positive associations have been reported between volume of tap water consumed and bladder cancer risk. This may be due to increased intake of carcinogenic chemicals contained in the water, such as arsenic (which is a recognised cause of bladder cancer) or disinfection by-products (e.g. trihalomethanes, the main by-product of chlorinated water), but the results of studies are not consistent.

Individuals with type II diabetes may have a modest increased risk of developing bladder cancer. A meta-analysis of studies investigating human papilloma virus (HPV) infection and bladder cancer estimated that risk could be increased by almost 3-fold in infected individuals. In another meta-analysis, bladder cancer risk was reduced by one-third in ever parous women (those who had had one or more children), and was elevated among those undergoing an early menopause. Non-smokers who regularly use non-steroidal anti-inflammatory drugs other than aspirin may have reduced bladder cancer risk, although aspirin use itself does not appear to be associated with risk. Those with higher levels of selenium measured in serum or toenails may also have lower risk. 

The geographical pattern of bladder cancer for both sexes was mainly determined by the pattern for men, due to their higher incidence rate (Map 11.1).

For men (Map 11.2), there were several areas of high relative risk, including the east coast from Louth to Wexford (including Dublin City), Donegal, and more scattered areas including parts of Kerry, Cork, North Down, Ards and north and east Belfast.

For women the geographical variation was less distinct (Map 11.3). As for men, there was an area of higher relative risk along the east coast, but there was a second area of increased risk which included most of Munster—a much larger area than for men. Donegal and parts of Belfast showed slightly increased levels of relative risk.

Map 11.1 Bladder cancer, smoothed relative risks: both sexes

Map 11.2 Bladder cancer, smoothed relative risks: males

Map 11.3 Bladder cancer, smoothed relative risks: females

The category of head and neck cancer incorporates cancers at 17 separate sites in the mouth, pharynx, larynx, middle ear and nasal sinuses (see Table 2.1). Head and neck cancer was the ninth most common cancer in Ireland, accounting for 1.6% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 4.0% in men (Table 12.1). The average number of new cases diagnosed each year was 170 in women and 438 in men. During 1995-2007, the number of new cases diagnosed increased by approximately 1% per annum.

The risk of developing head and neck cancer up to the age of 74 was 1 in 209 for women and 1 in 67 for men and was slightly higher in NI than in RoI for women. At the end of 2008, 568 women and 1,293 men aged under 65, and 600 women and 1,462 men aged 65 and over, were alive up to 15 years after their head and neck cancer diagnosis.

Table 12.1 Summary information for head and neck cancer in Ireland, 1995-2007

The age distribution for head and neck cancer showed that men tended to be diagnosed at an earlier age than women—68% of men were aged under 70 years, compared to 58% of women (Figure 12.1). Almost twice the percentage of women presented at 80 years and over compared to men (18% .v. 10%). The pattern of age at diagnosis was similar in RoI and in NI.

Figure 12.1 Age distribution of head and neck cancer cases in Ireland, 1995-2007, by sex

The incidence of head and neck cancer in women was low, with age-standardised rates ranging from 2.1 per 100,000 (in Japan) to 6.3 per 100,000 (in Demark) (Figure 12.2). Variation in rates was greater for men, with rates highest in Spain and Portugal, and lowest in Japan and New Zealand. Rates in RoI and NI were close to the median of the countries examined for men; NI ranked fifth highest of the 21 countries for women.

Figure 12.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: head and neck cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007) NOTE: Head and neck cancer was defined in GLOBOCAN 2008 by ICD10 codes C00-C14 and C32 but in this atlas we have used codes C01-C14 and C30-C32 (see Table 2.1). The incidence rates shown for NI and RoI in Figure 12.2 use the GLOBOCAN definition, and are consequently slightly inconsistent with data in the rest of this chapter.

Table 12.2 Risk factors for head and neck cancer, by direction of association and strength of evidence

More than 70% of head and neck cancers are considered to be due to tobacco and alcohol (Table 12.2; Hashibe et al., 2009). Tobacco smoking, and use of smokeless tobacco products, such as chewing tobacco or snuff, are causally related to cancer at many of the specific sites within this group. Risk increases with duration of smoking and number of cigarettes smoked, and falls with increasing time since quitting. Risk, particularly of laryngeal and pharyngeal cancers, is probably also increased in those who have never smoked themselves, but have a long duration of involuntary smoking (passive smoking) exposure at home or work. A causal relationship with alcohol intake is also clearly established. Compared to non- or occasional drinkers, light drinkers have a modest increased risk of oral and pharyngeal cancers, while the risk in heavy drinkers is increased by 5-fold for oral cancers and 7-fold for pharyngeal cancers (Turati et al., 2010).

Overall, individuals with one or more first-degree relatives affected with head and neck cancer have a modest raised risk of developing the disease themselves, although the combination of a positive family history and use of alcohol and tobacco confers a much higher risk. Risk of most head and neck cancers is higher in those of lower socio-economic status, probably reflecting social class variations in exposure to tobacco and, perhaps also, alcohol.

Evidence of infection with human papilloma viruses (HPV) has been found in the oral cavity and larynx. Moreover, sexual behaviours that have previously been associated with HPV infection, such as earlier age at sexual debut and more sexual partners, have also been associated with increased risk of head and neck cancer (Heck et al., 2010). Consequently, the International Agency for Research on Cancer has concluded that various strains of HPV are causally implicated in some head and neck cancers. Notably, HPV16 is considered a causal agent for cancers of the oral cavity, oropharynx and tonsil, and is likely to also be involved in the aetiology of laryngeal cancer. However, the natural history of oral HPV infection remains unclear.

Systematic reviews suggest that higher intake of fruit and vegetables (non-starchy or carotenoid-rich) may be associated with decreased risk of head and neck cancer. Similarly, risk of cancers of the oral cavity and pharynx, but not the larynx, may be lower in those who consume more coffee. Body leanness and being underweight have been associated with increased risk of head and neck cancer, and overweight and obesity with reduced risk, but the possibility of reverse causality cannot be excluded.

Due to the much higher incidence of head and neck cancer in men, the geographical pattern for both sexes (Map 12.1) was similar to that for men only.

Areas of high relative risk for men were scattered throughout the country—along the western seaboard, around Cork, Kerry, Tipperary South, Clare, Limerick, Dublin city and north Dublin, Belfast, Moyle, Larne, Cookstown, Derry, Newry and Mourne, Longford, Galway and Mayo (Map 12.2).

There was a quite different pattern of geographical variation for women with one large area of higher relative risk from north Leitrim/south Donegal to the east coast of NI, and covering most of the southern half of NI and the border counties in RoI. Other areas of higher risk were in Dublin city, north Dublin, the Inishowen peninsula in Donegal, the Dingle peninsula in Kerry and along the northern coast of NI from Derry to Moyle (Map 12.3).

Map 12.1 Head and neck cancer, smoothed relative risks: both sexes

Map 12.2 Head and neck cancer, smoothed relative risks: males

Map 12.3 Head and neck cancer, smoothed relative risks: females

Leukaemia describes a number of diseases of differing aetiology and clinical course. The four main subtypes are chronic lymphocytic leukaemia (CLL), chronic myeloid leukaemia (CML), acute lymphocytic leukaemia (ALL) and acute myeloid leukaemia (AML). The commonest of these (about 40% of all cases) is CLL (National Cancer Registry, 2010a). All leukaemias combined formed the tenth most common cancer in Ireland, accounting for 2.4% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 3.2% in men (Table 13.1). The average number of new cases diagnosed each year was 243 in women and 348 in men. During 1995-2007, the number of new cases diagnosed showed a small overall increase of 1% per annum. However, the numbers decreased slightly in NI, while increasing at approximately 2% per annum in RoI.

The risk of developing leukaemia up to the age of 74 was 1 in 170 for women and 1 in 98 for men and was higher in RoI than in NI. At the end of 2008, 847 women and 1,196 men aged under 65, and 663 women and 924 men aged 65 and over, were alive up to 15 years after their leukaemia diagnosis.

Table 13.1 Summary information for leukaemia in Ireland, 1995-2007

In contrast to most other cancers leukaemia was more common at the extremes of age—47% of women and 38% of men were aged either under 50 or over 79 years at diagnosis (Figure 13.1). Approximately one quarter of all new cases presented in the age group 70–79 years. The pattern of age at diagnosis was similar in RoI and in NI.

Figure 13.1 Age distribution of leukaemia cases in Ireland, 1995-2007, by sex

Leukaemia rates in RoI were relatively high compared with many other developed countries, while rates in NI were relatively low, especially for men (Figure 13.2). New Zealand, USA, Australia and Canada had the highest male rates, while Poland and Japan had the lowest. New Zealand and Italy had the highest female rates of the disease, while Poland and Japan also had the lowest female rates.

Figure 13.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: leukaemia
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 13.2 Risk factors for leukaemia, by direction of association and strength of evidence

The chronic leukaemias (CLL, CML) almost never affect children; AML may affect adults and children; while ALL is the most common type of leukaemia in children, but may affect adults also. The description of risk factors below refers to leukaemia in adults.

Ionizing radiation exposure has long been established as a risk factor for leukaemia. Risk of all types of leukaemia other than CLL is increased in individuals who receive particular types of radiation therapy or undergo multiple x-rays, or have protracted exposure to low-dose gamma radiation. A variety of anti-neoplastic drugs (alkylating agents), given as part of the treatment for other types of cancer, can cause leukaemia, most usually AML. Risk is increased in individuals with myelodysplastic syndrome, a diverse collection of conditions related to the blood and bone marrow which, if left untreated, may progress to AML. Infection with the human T-cell lymphotrophic virus, type 1 (HTLV-1) causes a rare type of leukaemia known as adult T-cell leukaemia. Approximately 10% of individuals with CLL report a family history of the condition or a related lymphoproliferative disorder, but the genes which account for increased susceptibility have not been identified.

Occupational exposure to benzene (an industrial solvent and precursor to basic industrial chemicals including drugs, plastics, synthetic rubber, and dyes); formaldehyde (used in the production of industrial resins that are then used in the manufacture of products such as adhesives and binders for wood products, plastic and synthetic fibres); and styrene (used in the production of polymers, which are incorporated into products such as rubber, plastic, insulation, fibreglass, and car parts) are all linked with leukaemia. Formaldehyde is particularly associated with AML. Workers in the rubber manufacturing industry have increased risk, but due to the complexity of exposures in this industry, the causative agents have not yet been firmly identified. Risk of myeloid leukaemias may also be raised in individuals employed in plants manufacturing pesticides.

In terms of lifestyle-related risk factors, tobacco smoking is a causal agent for myeloid leukaemia. Overweight and obesity may modestly increase risk of all four main subtypes, but the evidence-base is sparse.

Leukaemia had a very strong geographical pattern of relative risk, which was similar for men and women, although more pronounced in men (Maps 13.1-13.3).

The maps show a gradient of risk across the island, highest in the south-west and reducing uniformly across the country, with the lowest risk in the north-east (Map 13.1).

This pattern was more pronounced for men (Map 13.2) than women (Map 13.3). There was no marked variation in risk within the urban areas, which had similar risks to their surrounding rural areas.

Map 13.1 Leukaemia, smoothed relative risks: both sexes

Map 13.2 Leukaemia, smoothed relative risks: males

Map 13.3 Leukaemia, smoothed relative risks: females

Pancreatic cancer was the eleventh most common cancer in Ireland, accounting for 2.6% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 2.5% in men (Table 14.1). The average number of new cases diagnosed each year was 272 in women and 269 in men. During 1995-2007, the number of new cases diagnosed showed an overall increase of approximately 4% per annum.

The risk of developing pancreatic cancer up to the age of 74 was 1 in 169 for women and 1 in 123 for men and was similar in RoI and NI. At the end of 2008, 101 women and 115 men aged under 65, and 197 women and 154 men aged 65 and over, were alive up to 15 years after their pancreatic cancer diagnosis.

Table 14.1 Summary information for pancreatic cancer in Ireland, 1995-2007

Pancreatic cancer is a disease of older people; fewer than 20% of new cases were diagnosed in persons under 60 years old, while almost 60% presented at 70 years or over (Figure 14.1). The average age at diagnosis was older for women than men, with a similar age distribution in RoI and NI.

Figure 14.1 Age distribution of pancreatic cancer cases in Ireland, 1995-2007, by sex

Pancreatic cancer rates were highest in the Czech Republic for both men and women, and lowest in Sweden for men and in Portugal for women (Figure 14.2). Male rates of pancreatic cancer in both RoI and NI were close to the median of the countries shown; for women, they were below the median in NI and above the median in RoI.

Figure 14.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: pancreatic cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 14.2 Risk factors for pancreatic cancer, by direction of association and strength of evidence

Tobacco smoking is causally related to pancreatic cancer (Table 14.2) and accounts for 20-25% of cases (Maisonneuve & Lowenfels, 2010). Smokers of cigarettes, cigars and pipes are all at increased risk; risk increases with duration of smoking and cumulative smoking dose, and decreases to background levels 15 years after smoking cessation (Lynch et al., 2009). Use of smokeless tobacco is also a causal factor and there are suggestions that individuals exposed to environmental tobacco smoke either in childhood, or at home or work as adults, have increased risk of pancreatic cancer. Alcohol is also causally related to pancreatic cancer although the association appears to be restricted to individuals with high alcohol intake, and may be limited to men.

As regards medical conditions, around 5% of patients with long-term inflammation of the pancreas (chronic pancreatitis) are likely to develop pancreatic cancer over a 20 year period. Diabetics have increased risk, especially insulin users, but some uncertainties remain regarding the association (Magruder et al., 2011). In a few studies, insulin resistance and metabolic syndrome have also been linked to elevated pancreatic cancer risk, while use of metformin as a treatment for diabetes has been associated with reduced risk. Inverse associations have been reported between allergic conditions and atopy and pancreatic cancer risk.

Taller adults have increased risk, but height per se is unlikely to affect risk; it is most probably a marker for a risk factor related to linear growth in childhood. Risk of pancreatic cancer increases with increasing body fatness, by 6% for men and 12% for women per 5kg/m² increase in body mass index. In addition, most studies which have examined measures of abdominal fatness have reported positive findings. Some other lifestyle-related factors have been linked to pancreatic cancer (e.g. physical activity, aspects of diet), but the evidence regarding these remains somewhat uncertain.

An estimated 5-10% of pancreatic cancers arise as a result of genetic syndromes. One affected first degree relative confers a 75% increased risk.

Pancreatic cancer had a strong geographical pattern which was similar for men and women (Maps 14.1-14.3).

This consisted of a fairly smooth gradient, from the area of lowest risk in the north-east to the area of highest risk in the south-west (Map 14.1), highest around Cork city for men (Map 14.2) and centred on north Kerry in women, for whom there were also some areas of higher risk in Donegal (Map 14.3). The variation in risk was more pronounced for women than for men.

Map 14.1 Pancreatic cancer, smoothed relative risks: both sexes

Map 14.2 Pancreatic cancer, smoothed relative risks: males

Map 14.3 Pancreatic cancer, smoothed relative risks: females

Kidney cancer was the twelfth most common cancer in Ireland, accounting for 1.8% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 2.8% in men (Table 15.1). The average number of new cases diagnosed each year was 188 in women and 310 in men. During 1995-2007, the number of new cases diagnosed showed an overall increase of approximately 6% per annum in RoI; the rate of increase was lower in NI at approximately 3%.

The risk of developing kidney cancer up to the age of 74 was 1 in 188 for women and 1 in 98 for men and was similar in NI and RoI. At the end of 2008, 606 women and 959 men aged under 65, and 765 women and 1,070 men aged 65 and over, were alive up to 15 years after their kidney cancer diagnosis.

Table 15.1 Summary information for kidney cancer in Ireland, 1995-2007

Kidney cancer is a disease of older ages, with only about 30% of new cases diagnosed in persons under 60 years old, and the majority of new cases presenting between 60 and 79 years. A significant percentage of women (16%) were diagnosed at 80 years and over (Figure 15.1). Age at diagnosis was slightly younger in RoI than in NI.

Figure 15.1 Age distribution of kidney cancer cases in Ireland, 1995-2007, by sex

Kidney cancer rates were highest in the Czech Republic, followed by the USA, for both men and women (Figure 15.2). The lowest rates occurred in Japan and Portugal for both men and women. Compared to other developed countries, the rates of kidney cancer were slightly below the median in both RoI and NI.

Figure 15.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: kidney cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007) NOTE: KIdney cancer was defined in GLOBOCAN 2008 by ICD10 codes C64-C66 but in this atlas we have used codes C64-C65 (see Table 2.1.). The incidence rates shown for NI and RoI in Figure 15.2 use the GLOBOCAN definition, and are consequently slightly inconsistent with data in the rest of this chapter.

Table 15.2 Risk factors for kidney cancer, by direction of association and strength of evidence

The two most common types of kidney cancer are renal cell carcinoma and urothelial cell carcinoma of the renal pelvis; the majority (over 90%) are renal cell carcinomas. Tobacco smoking is causally associated with renal cell cancer in both men and women (Table 15.2): there is a dose-response relationship with number of cigarettes smoked and, in some studies, risk reduces after cessation of smoking (International Agency for Research on Cancer, 2004b). An association between use of smokeless tobacco and kidney cancers in general has also been suggested.

As regards other lifestyle-related risk factors, pooled analysis of prospective studies suggests that drinking alcohol may be associated with a moderately decreased risk of renal cell cancers. In contrast, studies consistently show that body fatness is associated with elevated risk. Overall, risk increases by almost one-third per 5kg/m² increase in body mass index and perhaps to a slightly greater extent in women than men.

Advanced kidney disease, which makes dialysis necessary, raises the risk of renal cell cancer, but the underlying mechanism is unclear. Hypertension has also been associated with a 60% increased risk of renal cell cancer. Risk may also be raised in those who have used antihypertensive drugs, but whether the associations hold in men and women and after adjustment for other risk factors is somewhat uncertain.

Exposure to cadmium and cadmium compounds, which generally takes place occupationally, and to arsenic or inorganic arsenic compounds, which may be occupational or ingested in food or drinking water, is likely to cause kidney cancer. Occupational exposure to tricholoroethylene, a hydrocarbon commonly used as an industrial solvent, has been associated with a modestly raised risk, but concerns remain about whether the association could be due to uncontrolled confounding.

After accounting for other risk factors, risk of kidney cancer in general, and renal cell cancer in particular, is increased by between 40% and 100% in those with at least one affected first degree relative.

The geographical pattern of relative risk for kidney cancer showed only modest variation, which was different for men and women (Maps 15.1-15.3).

For both sexes combined, the area of highest relative risk was mainly confined to Leinster, with lower relative risks in the western half of Ireland (Map 15.1).

For men, the relative risk was higher in a wide band across the country extending from Sligo to Wexford. The relative risks were lower in the south-west and north (Map 15.2).

For women, there were just two areas of higher relative risk, along the north coast (from Donegal in RoI to Moyle/Larne in NI) and in the area from Louth to Wicklow, while the west had lower relative risks (Map 15.3).

Map 15.1 Kidney cancer, smoothed relative risks: both sexes

Map 15.2 Kidney cancer, smoothed relative risks: males

Map 15.3 Kidney cancer, smoothed relative risks: females

Oesophageal cancer was the thirteenth most common cancer in Ireland, accounting for 1.8% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 2.7% in men (Table 16.1). The average number of new cases diagnosed each year was 182 in women and 301 in men. During 1995-2007, the number of new cases diagnosed increased by approximately 2% per annum.

The risk of developing oesophageal cancer up to the age of 74 was 1 in 258 for women and 1 in 105 for men and was similar in NI and RoI. At the end of 2008, 118 women and 278 men aged under 65, and 308 women and 423 men aged 65 and over, were alive up to 15 years after their oesophageal cancer diagnosis.

Table 16.1 Summary information for oesophageal cancer in Ireland, 1995-2007

The age distribution at diagnosis was different for men and women (Figure 16.1). More than half of men, but only one-third of women, presented at under 70 years of age, while a further third of women, but only 16% of men, was aged 80 years or older at diagnosis. The pattern was similar in RoI and NI.

Figure 16.1 Age distribution of oesophageal cancer cases in Ireland, 1995-2007, by sex

Among developed countries, only the UK had higher rates of oesophageal cancer in women than NI and RoI (Figure 16.2). Among men, NI and RoI also had relatively high rates of the disease, with only UK, Netherlands and Japan having higher rates. The lowest rates were in Sweden and Italy for men and in Austria and Portugal for women.

Figure 16.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: oesophageal cancer
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 16.2 Risk factors for oesophageal cancer, by direction of association and strength of evidence

The two main types of oesophageal cancer are squamous cell carcinoma and adenocarcinoma. Some risk factors are shared by both types, while others are involved in one type only. Tobacco smoking causes both squamous cell carcinoma and adenocarcinoma of the oesophagus (Table 16.2). Smokers have at least a two-fold higher risk than non-smokers and risk increases with number of cigarettes smoked daily and duration of smoking. Use of smokeless tobacco products (e.g. snuff, chewing tobacco) is also associated with increased disease risk. Alcohol is also causally related to oesophageal cancer and risk increases with amount consumed.

Obesity and overweight are positively associated with adenocarcinoma. In contrast, higher levels of body fatness are either unrelated to risk of squamous cell carcinoma, or associated with a decreased risk (Smith et al., 2008). A history of gastro-oesophageal reflux disease has been associated with increased risk of adenocarcinoma, but not of squamous cell carcinoma. Gastric atrophy may be positively associated with squamous cell carcinomas and negatively associated with adenocarcinoma. Infection with the Helicobacter pylori (H pylori) bacterium has been associated with reduced risk of adenocarcinoma, while infection with human papilloma virus may play a role in squamous cell carcinoma.

Various aspects of diet have been linked with oesophageal cancer risk. Higher intakes of fruit and vegetables, particularly those containing beta-carotene (yellow, orange and green fruits and green leafy vegetables) or vitamin C, probably reduce risk. Higher intakes of red or processed meat may increase risk, but the evidence is less consistent than for fruit and vegetables. Risk may also be increased in those with higher intakes of pickled vegetables, and those who prefer to consume their hot drinks at a high temperature.

Risk of oesophageal cancer is higher in those of low socio-economic status, probably reflecting variations in exposure to tobacco and other lifestyle risk factors by social class.

Map 16.1 Oesophageal cancer, smoothed relative risks: both sexes

Map 16.2 Oesophageal cancer, smoothed relative risks: males

Map 16.3 Oesophageal cancer, smoothed relative risks: females

Ovarian cancer was the fourth most common cancer for women in Ireland, accounting for 4.6% of all malignant neoplasms, excluding non-melanoma skin cancer (Table 17.1). The average number of new cases diagnosed each year was 479. During 1995-2007, there was little change in the number of new cases diagnosed per year in RoI while the numbers decreased slightly in NI.

The risk of developing ovarian cancer up to the age of 74 was 1 in 71 and was similar in NI and RoI. At the end of 2008, 1,441 women aged under 65 and 994 women aged 65 and over were alive up to 15 years after their cancer diagnosis.

Table 17.1 Summary information for ovarian cancer in Ireland, 1995-2007

The age distribution of ovarian cancer at diagnosis was more uniform than for most other cancers (Figure 17.1). Approximately 18% of cases presented under 50 years of age, 21% between 50 and 59, 25% between 60 and 69 and 23% between 70 and 79; the remainder (13%) presented at 80 years or over. There was a similar age pattern in RoI and NI.

Figure 17.1 Age distribution of ovarian cancer cases in Ireland, 1995-2007

In 2008, estimated rates of ovarian cancer were high in the UK, RoI and NI and also in Poland and the Czech Republic (Figure 17.2). Rates of the disease were lowest in Portugal, followed by Japan, Australia and Canada. Differences in the coding of borderline ovarian cancers by cancer registries may account for some of the international differences in incidence rate.

Figure 17.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: ovarian cancer

Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 17.2 Risk factors for ovarian cancer, by direction of association and strength of evidence

Family history of the disease is the most important risk factor for ovarian cancer. Women with at least one affected first degree relative have a three-fold increased risk of developing ovarian cancer themselves. Together with breast cancer, ovarian cancer is a component of several autosomal dominant cancer syndromes, most notably the BRCA1 and BRCA2 mutation syndromes. Women who have mutations in the BRCA1 or BRCA2 genes have a high chance of developing ovarian cancer over their lifetime (39% risk for BRCA1 mutation carriers and 11% for BRCA2 mutation carriers up to the age of 70; Antoniou et al., 2003).

Endogenous and exogenous oestrogens play a key role in the aetiology of ovarian cancer. Not having children (nulliparity) is associated with increased risk and, by contrast, each additional child confers a 20% reduction in risk (Riman et al., 2004). Women who have used fertility drugs may also have raised risk, especially if they remain nulligravid. Risk is reduced in women who have had bilateral tubal ligation (their fallopian tubes “tied”) or hysterectomy with ovarian conservation. In terms of exogenous hormones, use of hormone replacement therapy, particularly oestrogen-only formulations, is clearly related to increased risk, although formulations which include progestin are associated with only a modest risk. Women who have used combined oestrogen-progestogen oral contraceptives have decreased risk. Protection increases with increasing duration of use, and the reduction in risk persists for more than 30 years after use has ceased. Breastfeeding has also been associated with lower risk of ovarian cancer, but uncertainties remain in the evidence.

Risk of ovarian cancer is raised in taller women, but tallness in itself is unlikely to be a causal factor; it is most likely a marker for factors relating to promoters of growth in childhood (World Cancer Research Fund / American Institute of Cancer Research, 2007).

Several lifestyle factors have been related to ovarian cancer risk, but most of the associations remain somewhat uncertain. The exception is smoking, which is causally related to mucinous ovarian tumours, but not to other tumour types (Jordan et al., 2006). Higher intake of non-starchy vegetables may be associated with reduced risk, but this category includes a wide and disparate group of vegetables with many different plant food constituents (e.g. fibre, folate, flavonoids) and the contributions of the different constituents have not been unravelled (World Cancer Research Fund / American Institute of Cancer Research, 2007). There may be a modest (approximately 20%) reduction in risk among women with the highest, versus those with the lowest, levels of recreational physical activity. Body fatness and/or obesity may also be related to increased risk.

Map 17.1 Ovarian cancer, smoothed relative risks

The group of cancers of the brain and other central nervous system includes those which occur in the brain (more than 90% of the cancers), meninges (~3%) and cranial nerves or spinal cord (~3%); (National Cancer Registry, 2010b). Brain and other central nervous system cancer (CNS) was the fifteenth most common cancer in Ireland, accounting for 1.7% of all malignant neoplasms, excluding non-melanoma skin cancer, in women and 2.1% in men (Table 18.1). The average number of new cases diagnosed each year was 174 in women and 234 in men. During 1995-2007, the number of new cases diagnosed showed an overall increase of approximately 3% per annum.

The risk of developing brain and CNS cancer before the age of 75 was 1 in 204 for women and 1 in 134 for men and was slightly higher in RoI than in NI. At the end of 2008, 563 women and 706 men aged under 65, and 123 women and 111 men aged 65 and over, were alive up to 15 years after their cancer diagnosis.

Table 18.1 Summary information for brain and other central nervous system cancer in Ireland, 1995-2007

Cancer of the brain and CNS was predominantly a disease of younger persons (Figure 18.1). Approximately half of all new cases presented under 60 years of age (55% of men and 47% of women). The average age at diagnosis was younger in NI than in RoI.

Figure 18.1 Age distribution of brain and other central nervous system cancer cases in Ireland, 1995-2007, by sex

Denmark, Norway and Sweden had the highest rates of brain and CNS cancer in 2008 for both men and women, in a comparison of developed countries (Figure 18.2). Japan, Russia and Austria had the lowest rates of the disease. Both RoI and NI had slightly higher than median rates for men, as did RoI for women.

Figure 18.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: brain and other central nervous system cancers
females	males
Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 18.2 Risk factors for brain and other central nervous system cancer, by direction of association and strength of evidence

Most brain tumours in adults start in glial cells and are known as gliomas. Among the gliomas, astrocytomas are most common. Although cancers of the brain and other CNS may occur in children, most cases present in adults and the following description of risk factors relates only to cancers in adults.

Little is known about the aetiology of brain and other CNS cancers, in part because it can be difficult to obtain reliable exposure data from affected individuals, and the cancers are relatively rare, thus limiting prospective (cohort) studies. The most firmly established risk factor is exposure to high doses of ionizing radiation, usually from x-rays or radiation therapy. Whether exposure to radio-frequency electromagnetic fields (RF-EMF), such as those emitted by mobile phones, mobile phone base stations and broadcast antennas, causes cancers of the brain is a matter of considerable controversy. The International Agency for Research on Cancer recently concluded that there was limited evidence on carcinogenicity of RF-EMF, based on associations between wireless phones and glioma (and acoustic neuroma, which is a benign tumour of the nerve which connects the ear to the brain), but some members of the Expert Group considered the evidence-base to be inadequate (Baan et al., 2011).

Risk of gliomas and oligodendroglial tumours is reduced in individuals with asthma or allergic conditions, such as eczema and hay fever. This may also hold for meningiomas, but few studies have been done.

Farming is related to a modest increased risk of cancer of the brain, but the specific exposures which may increase risk have not been identified. Individuals occupationally exposed to formaldehyde—which is used in the production of industrial resins that are then used in the manufacture of products such as adhesives and binders for wood products, plastic and synthetic fibres—may have moderately increased risk of brain and other CNS cancers, but the evidence is not consistent.

Map 18.1 Brain and other central nervous system cancer, smoothed relative risks: both sexes

Map 18.2 Brain and other central nervous system cancer, smoothed relative risks: males

Map 18.3 Brain and other central nervous system cancer, smoothed relative risks: females

Cancer of the corpus uteri (uterine cancer) was the sixth most common cancer in women in Ireland, accounting for 3.9% of all malignant neoplasms, excluding non-melanoma skin cancer, in women (Table 19.1). The average number of new cases diagnosed each year was 403. During 1995-2007, the number of new cases diagnosed per year increased by 7% in NI and 3% in RoI.

The risk of developing uterine cancer up to the age of 74 was 1 in 77 and was slightly higher in NI than in RoI. At the end of 2008, 1,756 women aged under 65, and 2,419 aged 65 and over, were alive up to 15 years after diagnosis.

Table 19.1 Summary information for uterine cancer in Ireland, 1995-2007

Almost 80% of cases of uterine cancer were diagnosed between 50 and 79 years of age, with 27% presenting between 50 and 59 years, 30% between 60 and 69 years and 22% between 70 and 79 years (Figure 19.1). Only 11% of cases were diagnosed in women aged less than 50, with a further 10% diagnosed in those aged 80 and older.

Figure 19.1 Age distribution of cases of uterine cancer in Ireland, 1995-2007

Rates of uterine cancer varied considerably among developed countries, with rates highest in the Czech Republic and Norway and lowest in Japan and Portugal (Figure 19.2). In RoI the incidence rate of the disease was low compared to other countries, while in NI the rate was slightly higher than the median. Variation between countries in the percentage of cases assigned to “uterus, part unspecified” (which are not included in the data below) may account for some of the international variation.

Figure 19.2 Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: uterine cancer

Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 19.2 Risk factors for uterine cancer, by direction of association and strength of evidence

Hormones play a major role in the aetiology of uterine cancer. A surge in incidence of these cancers in the USA in the late-1960s and mid-1970s led to the discovery of the role of hormone replacement therapy (Jick et al., 1980). Oestrogen-only formulations of hormone replacement therapy are now considered to be a causal agent and risk is also increased in women who use regimes which contain both oestrogen and progestogen, but in which the progestogen is taken for less than 15 days per month. Tamoxifen, a selective oestrogen receptor modulator used in the prevention and treatment of breast cancer, also causes uterine cancer. In contrast, use of combined oestrogen-progestogen oral contraceptives is protective. Risk is reduced by about half in women who have taken oral contraceptives, and the protective effect lasts for around two decades after cessation of use. However, it should be noted that these findings have generally been based on older high-dose oral contraceptives. In addition, reproductive factors associated with increased exposure to endogenous oestrogens may also influence risk. Women who have not had children (nulliparous) are at increased risk, while risk reduces with increasing number of full-term pregnancies. Risk may be raised in women with early menarche and late menopause, while breastfeeding may be protective. The strong and consistent association between body fatness and uterine cancer (60% increase in risk per 5 kg/m² increase in body mass index) may also operate through effects on oestrogens and other hormones, such as insulin (World Cancer Research Fund / American Institute of Cancer Research, 2007). These effects are also a possible explanation for the observed relationship between higher levels of physical activity and reduced disease risk. As regards insulin levels more directly, there is a two to three-fold increased risk in diabetic individuals (both type 1 and type 2 diabetes).

Women who have one or more first-degree relatives affected by uterine cancer have a two-fold increased risk of developing the disease. Part of this is due to women in families affected by hereditary nonpolyposis colorectal cancer (HNPCC) syndrome, also known as Lynch syndrome. Depending on which genetic mutations they carry, these women have a 20-55% chance of developing uterine cancer in their lifetime.

Map 19.1 Cancer of the corpus uteri, smoothed relative risks

Cancer of the uterus is classified by ICD-10 into three sites—cancer of the cervix uteri (cervical cancer; discussed in this chapter), cancer of the corpus uteri (uterine cancer; see Chapter 19) and cancer of the uterus, part unspecified. “Part unspecified” cases make up less than 5% of all cancers of the uterus and are not considered in this atlas.

Cancer of the cervix uteri was the eighth most common cancer for women in Ireland, accounting for 2.8% of all malignant neoplasms, excluding non-melanoma skin cancer, in women (Table 20.1). The average number of new cases diagnosed each year was 289. During 1995-2007, there was an increase of 5% in the number of new cases diagnosed per year in RoI, while the numbers remained fairly constant in NI.

The risk of developing cervical cancer up to the age of 74 was 1 in 124 and was slightly higher in RoI than in NI. At the end of 2008, 2,484 women aged under 65 and 418 aged 65 and over were alive up to 15 years after their diagnosis.

Table 20.1 Summary information for cervical cancer in Ireland, 1995-2007

Cervical cancer was predominantly a disease of younger women (Figure 20.1). Almost 60% of new cases presented in those aged less than 50, and over three-quarters under 60. The pattern was similar in RoI and NI.

Figure 20.1 Age distribution of cases of cervical cancer in Ireland, 1995-2007

Cervical cancer incidence rates in 2008 were highest in the Czech Republic and Russia and lowest in Australia and New Zealand (Figure 20.2). Rates of the disease in RoI were slightly higher than the median, while in NI the rates was close to the median. Variation between countries in the percentage of cases assigned to “uterus, part unspecified” (which are not included in the data below) may account for some of the international variation.

Figure 20.2: Estimated incidence rate per 100,000 in 2008 for selected developed countries compared to 2005-2007 incidence rate for RoI and NI: cervical cancer

Source: GLOBOCAN 2008 (Ferlay et al., 2008) (excluding RoI and NI data, which is derived from Cancer Registry data for 2005-2007)

Table 20.2 Risk factors for cervical cancer, by direction of association and strength of evidence

Many strains of human papilloma viruses (HPV) infect the genital squamous epithelia. Some strains (known as "low-risk") cause genital warts while other strains (known as "high-risk") cause cervical cancer. The association between cervical cancer and these high-risk types of HPV infection is so strong that HPV is considered to be a necessary cause of the disease (Bosch et al., 2002). Infection with high-risk HPV is very common, and most women who have been sexually active will be infected at some time during their lifetime (Bosch et al., 2008). In most women infection causes no symptoms and clears naturally within a few months. However, some women become re-infected and the virus persists; susceptibility to persistent infections is thought to increase risk of developing cervical lesions. The factor most consistently associated with risk of genital HPV infection is number of sexual partners (Winer and Koutsky, 2004).

Infection with human immunodeficiency virus, type 1 (HIV-1) is also recognised to cause cervical cancer.

As regards other risk factors, there is a causal relationship between smoking and squamous cell cancer of the cervix, which persists after adjustment for HPV infection. In the relatively few studies of adenocarcinoma and adeno-squamous cell carcinoma, no relationship with smoking has been found (International Agency for Research on Cancer, 2004b). Cervical cancer risk is raised in women who have used combined oestrogen-progestogen oral contraceptives for at least five years. Risk falls with increasing time since last use, and after 10 years returns to background levels. Risk also increases with the number of children that a woman has had (parity).

Women of lower socio-economic status have raised cervical cancer risk. While partly a function of variations in exposure to risk factors (de Sanjosé et al., 1997), this also reflects social class differences in access to cervical smear tests or participation in organised screening programmes (Segnan, 1997).